To the Editor:
In individuals with proteinuric kidney disease, the relative content of albumin to other lower molecular weight (LMW) proteins may provide information on the pathological process responsible for proteinuria. Tubular lesions or excessive systemic production could account for a relative predominance of LMW proteins that normally pass across the glomerular filtration barrier and then undergo catabolism.1 By contrast, lesions in the glomerular filtration barrier may allow passage of albumin in addition to LMW proteins, with albumin predominating due to the much higher concentration of albumin in plasma.2 The urine albumin-to-protein ratio (UAPR) has been proposed as a potential aid in differential diagnosis, but its ability to identify histopathologic lesions and clinicopathologic diagnoses, such as tubulointerstitial lesions and disease, has not been studied in depth.3,4
We evaluated UAPR, urine albumin-to-creatinine ratio (UACR), and urine protein-to-creatinine ratio (UPCR) in 338 individuals from the Boston Kidney Biopsy Cohort (BKBC) Study, which enrolled adults undergoing clinically-indicated native kidney biopsy between September 2006 and June 2016.5 Kidney biopsies were adjudicated under light microscopy by two experienced kidney pathologists who provided semiquantitative scores of kidney inflammation, fibrosis, vascular sclerosis, and tubular injury (Table S1). All participants’ charts were reviewed alongside the histopathologic evaluations to provide the final primary clinicopathologic diagnosis. We calculated UAPR from measurements of spot urine albumin and urine protein in samples collected on the day of biopsy. Multivariable linear regression models tested the association of UAPR with histopathologic lesions. We calculated the area under the receiver operating characteristic curve (AUROC) to compare the ability of UAPR, UACR, and UPCR to distinguish tubulointerstitial disease from other forms of kidney disease. We identified the threshold of UAPR that maximized sensitivity and specificity to differentiate tubulointerstitial disease from other forms of kidney disease using the Youden index.6 Full methods are in Item S1.
Baseline characteristics of BKBC participants by tertiles of UAPR are provided in Table S2. Mean age was 53.0 ± 16.5 years, mean eGFR was 54.7 ± 34.9 ml/min/1.73m2, median UACR was 1.0 [0.3 – 2.8] g/g creatinine, and median UPCR was 1.62 [0.6 – 3.7] g/g creatinine. The median UAPR was 0.69 [0.55 – 0.77]. UACR and UPCR were highly correlated (rs = 0.97, P<0.001). The most common primary clinicopathologic diagnoses were proliferative glomerulonephritis (32%), non-proliferative glomerulopathy (18%), diabetic nephropathy (15%), advanced glomerulosclerosis (10%), and tubulointerstitial diseases (9%) (Tables S2 and S3). In patients with tubulointerstitial diseases, the median UAPR was 0.24 [0.14 – 0.54], which was the lowest among patients with other clinicopathologic diagnoses (Table S4).
In models adjusted for eGFR and comparing against the most frequent diagnosis of proliferative glomerulonephritis, UAPR was significantly lower in those with tubulointerstitial disease and higher in those with non-proliferative glomerulopathy, paraprotein-related disease, and diabetic nephropathy (Figure 1A). In analyses of histopathology adjusting for eGFR, UAPR was significantly higher in patients with greater mesangial expansion, global glomerulosclerosis, segmental sclerosis, and interstitial fibrosis/tubular atrophy, and lower in those with greater inflammation in the preserved interstitium (Figure 1B).
Figure 1. Adjusted differences in UAPR by clinicopathologic diagnostic categories and histopathologic lesions.
A) Models were fitted using UAPR as the outcome and primary clinicopathologic diagnosis as the predictor variable in the linear regression model.
Reference is proliferative glomerulonephritis. Model is adjusted for age, sex, race, and eGFR.
B) Models were fitted using UAPR as the outcome and each histopathologic lesion as the predictor variable in linear regression models. Each model is adjusted for age, sex, race, and eGFR.
Reference is absence of lesion for glomerular inflammation, segmental sclerosis, inflammation in the non-fibrosed interstitium; Reference is none/mild lesion severity for mesangial expansion, acute tubular injury, arterial sclerosis, arteriolar sclerosis; Reference is 0–25% of cortical volume affected for global glomerulosclerosis, inflammation in the fibrosed interstitium, interstitial fibrosis/tubular atrophy.
Abbreviations: UAPR, urine albumin to protein ratio; IFTA, interstitial fibrosis and tubular atrophy; ATI, acute tubular injury
The AUROC for UAPR to distinguish tubulointerstitial disease from other forms of kidney disease was 0.84 (95% CI 0.74 – 0.93); for UACR and UPCR, the AUROCs were 0.76 (95% CI 0.65 – 0.86) and 0.66 (95% CI 0.55 – 0.76). The AUROC of UAPR was statistically significantly higher than those of UACR and UPCR (Figure 2A). The threshold of UAPR that maximized sensitivity and specificity was ≤0.54. This threshold had 78.1% sensitivity, 82.0% specificity, 30.4% positive predictive value, and 97% negative predictive value (Table S5). The addition of eGFR improved the ability to identify tubulointerstitial disease compared to other forms of kidney disease for all 3 spot urine protein ratios (Figure 2B). In subgroup analyses of patients with <3 grams of proteinuria and eGFR <60 ml/min/1.73m2, the results were similar (Figure 2C–D). Similar results were also seen in patients with ‘abnormal eGFR’ as the reason for biopsy (Figures 2E–F).
Figure 2. Receiver operating characteristic curves of spot urine protein ratios for the diagnosis of tubulointerstitial disease compared to other forms of kidney disease.
A) All participants.
B) All participants. Model additionally includes eGFR.
C) Participants with < 3 g/g proteinuria and eGFR < 60 ml/min/1.73m2
D) Participants with < 3 g/g proteinuria and eGFR < 60 ml/min/1.73m2. Model additionally includes eGFR.
E) Participants with the reason for kidney biopsy as ‘abnormal eGFR’
F) Participants with the reason for kidney biopsy as ‘abnormal eGFR’. Model additionally includes eGFR.
Several limitations warrant consideration. First, we excluded patients with proteinuria <100 mg/g due to weak correlation between UACR and UPCR at such low levels, as previously reported.7–10 Second, we did not have information on the presumed diagnosis (e.g., tubulointerstitial vs. glomerular processes) prior to biopsy. Finally, the number of individuals within each clinicopathologic diagnostic category was relatively small, particularly for rare diseases.
In summary, we found that lower UAPR was more commonly found in biopsy-confirmed tubulointerstitial disease compared to other forms of kidney disease involving the glomeruli, consistent with our understanding of the origins of proteinuria. A ratio below 0.54 had approximately 80% sensitivity and specificity for the diagnosis of tubulointerstitial disease; in this cohort of biopsy-confirmed and primarily glomerular kidney diseases, UAPR had relatively low PPV but high NPV. Accordingly, lower UAPR could aid in the differential diagnosis of tubulointerstitial compared to glomerular diseases but is unlikely to obviate the need for kidney biopsy in proteinuric kidney disease.
Supplementary Material
Supplementary Material Table of Contents
Item S1
Tables and Figures:
Table S1. Histopathologic scoring system for light microscopy in the BKBC Study
Table S2. Baseline Characteristics of BKBC Study participants
Table S3. Clinicopathologic diagnoses in the BKBC Study
Table S4. Heatmap of eGFR, UPCR, UACR and UAPR according to clinicopathologic diagnostic categories
Table S5. Performance characteristics of UAPR to diagnose tubulointerstitial disease compared to other forms of kidney disease
Acknowledgements
We thank the members of the laboratory of S.S.W. for their invaluable assistance in the Boston Kidney Biopsy Cohort. This work was conducted with support from Harvard Catalyst. The Harvard Clinical and Translational Science Center (National Center for Advancing Translational Sciences, National Institutes of Health Award UL1TR001102) and financial contributions from Harvard University and its affiliated academic healthcare centers. The content is solely the responsibility of the authors and does not necessarily represent the official views of Harvard Catalyst, Harvard University and its affiliated academic healthcare centers, or the NIH. The funders of this study did not have any role in the study design, collection, analysis, interpretation of the data, writing the report, or decision to submit the manuscript for publication.
Funding
This work was supported by National Institutes of Health (NIH) grant R01DK093574. AS is supported by NIH grants K23DK120811, U01AI163081, Kidney Precision Medicine Project Opportunity Pool grant under award U2CDK114886, and the American Society of Nephrology Carl W. Gottschalk Award. JL is funded by the Innovation Seedling Project (No. 2021029), Science and Technology Department of Sichuan Province, China.
Footnotes
Disclosures
AS reports personal fees from Horizon Therapeutics PLC, CVS Caremark, AstraZeneca, Bayer AG, FNIH, and Tate & Latham (medicolegal consulting). SSW reports the following: Employer: Boston Medical Center; Consultancy: Wolters Kluewer, GSK, Ikena, Senda Biosciences, Strataca/3ive, BioMarin, Regeneron, Bain, CANbridge; Ownership Interest: Amazon, Apple, Southwest Airlines, Boeing, Walt Disney, Citigroup; Research Funding: Vertex, Pfizer, JNJ; and Other Interests or Relationships: expert witness for litigation related to lab testing (Davita), PPIs (Pfizer), PFAO exposure (Dechert LLP). The remaining authors declare that they have no relevant financial interests.
References
- 1.Guder WG, Ivandic M, Hofmann W. Physiopathology of proteinuria and laboratory diagnostic strategy based on single protein analysis. Clin Chem Lab Med. 1998;36(12):935–939. [DOI] [PubMed] [Google Scholar]
- 2.Rennke HG, Olson JL, Venkatachalam MA. Glomerular filtration of macromolecules: normal mechanisms and the pathogenesis of proteinuria. Contrib Nephrol. 1981;24:30–41. [DOI] [PubMed] [Google Scholar]
- 3.Smith ER, Cai MM, McMahon LP, Wright DA, Holt SG. The value of simultaneous measurements of urinary albumin and total protein in proteinuric patients. Nephrol Dial Transplant. 2012;27(4):1534–1541. [DOI] [PubMed] [Google Scholar]
- 4.Ohisa N, Kanemitsu K, Matsuki R, Suzuki H, Miura H, Ohisa Y, et al. Evaluation of hematuria using the urinary albumin-to-total-protein ratio to differentiate glomerular and nonglomerular bleeding. Clin Exp Nephrol. 2007;11(1):61–65. [DOI] [PubMed] [Google Scholar]
- 5.Srivastava A, Palsson R, Kaze A, Chen M, Palacios P, Sabbisetti V, et al. The Prognostic Value of Histopathologic Lesions in Native Kidney Biopsy Specimens: Results from the Boston Kidney Biopsy Cohort Study. J Am Soc Nephrol. 2018;29(8):2213–2224. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Ruopp MD, Perkins NJ, Whitcomb BW, Schisterman EF. Youden Index and optimal cut-point estimated from observations affected by a lower limit of detection. Biom J. 2008;50(3):419–430. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Sumida K, Nadkarni GN, Grams ME, Sang Y, Ballew SH, Coresh J, et al. Conversion of Urine Protein-Creatinine Ratio or Urine Dipstick Protein to Urine Albumin-Creatinine Ratio for Use in Chronic Kidney Disease Screening and Prognosis : An Individual Participant-Based Meta-analysis. Ann Intern Med. 2020;173(6):426–435. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Kim SM, Lee CH, Lee JP, Oh YK, Kim YS, Kim S, et al. The association between albumin to creatinine ratio and total protein to creatinine ratio in patients with chronic kidney disease. Clin Nephrol. 2012;78(5):346–352. [DOI] [PubMed] [Google Scholar]
- 9.Fisher H, Hsu CY, Vittinghoff E, Lin F, Bansal N. Comparison of associations of urine protein-creatinine ratio versus albumin-creatinine ratio with complications of CKD: a cross-sectional analysis. Am J Kidney Dis. 2013;62(6):1102–1108. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Weaver RG, James MT, Ravani P, Weaver CGW, Lamb EJ, Tonelli M, et al. Estimating Urine Albumin-to-Creatinine Ratio from Protein-to-Creatinine Ratio: Development of Equations using Same-Day Measurements. J Am Soc Nephrol. 2020;31(3):591–601. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.