Urine Protein Measurement Issues Greg Miller, PhD Professor of Pathology Virginia Commonwealth University Richmond, VA (USA)
Jan 15, 2016
Urine Protein Measurement Issues
Greg Miller, PhD
Professor of Pathology
Virginia Commonwealth University
Richmond, VA (USA)
Outline
• Albumin molecules in urine
• Albumin measurement
• Total protein measurement
Albumin information is based on a
conference in March 2007 to address
standardization issues for urine
albumin/creatinine measurement and
reporting
Molecular forms of albumin
• Conformation is influenced by ligand binding
• Urine concentrates many ligands
• Unpaired cysteine-34 can form albumin dimers and other covalent modifications (plasma or urine)
• 1-10% is glycated in plasma (higher in diabetes)
• Glycated forms are a greater proportion in urine - attributed to differential uptake in tubules
• Large (>5 kDa) and small (500-5000 Da) fragments have been identified in plasma and urine
• C and N terminal truncation occurs
Albumin molecules in urine
• Conformation changes in plasma may influence the filtration rate at the glomerulus
• Fragmented forms in plasma should be more easily filtered by the glomerulus
• Tubular uptake is receptor mediated and may influence enrichment of modified forms in urine
• Proteases and chemical modifications occur in the urinary tract and in the urine after collection
• Influence of pH, osmolality, contact with sediment, adsorption to containers, and other sample handling factors are not well understood
Outline
• Albumin molecules in urine
• Albumin measurement
• Total protein measurement
Albumin measurement procedures
• Immunoassay
• Influenced by the albumin epitope(s) recognized by the antibody, and by reactivity with modified forms of albumin
• Evidence that polyclonal immunoassays are reactive with some modified forms
• HPLC
• Albumin may not be resolved from other urine proteins (causing overestimation)
• Hypothesis of non-immunoreactive albumin may be related to non-specificity
Measurement issues
• Influence of urine matrix variability and molecular forms of albumin are not well understood
• No urine albumin reference material
• No reference measurement procedure
• Calibration traceability is to diluted CRM 470 (ERM DA470) serum protein reference material
• Details of traceability design, dilution protocols, and measurement implementation are not standardized and appear to influence calibration uniformity between methods
• Some methods use the molar absorptivity of albumin in solution for calibration
Current status of albumin measurement
Proficiency Testing suggested:
• A range of results for the same sample
• Influenced by non-commutability of samples
• Urine dipstick results were highly variable
• A range of imprecision; with some methods having acceptable imprecision
• A variety of reporting units for albumin concentration, excretion rate, and albumin/creatinine ratio
PT example: CAP urine albumin (USA)
pooled human urine supplemented with albumin, creatinine and other substances, liquid, within method comparison
0
5
10
15
20
25
30
0 20 40 60 80 100
Peer group mean, mg/L
Am
on
g la
bo
rato
ry C
V, %
Inst. 1 (N=11)
Inst. 2 (N=79)
Inst. 3 (N=69)
Inst. 4 (N=194)
Inst. 5 (N=59)
Inst. 6 (N=59)
Inst. 7 (N=44)
Inst. 8 (N=39)
Inst. 9 (N=207)
Inst. 10 (N=112)
Inst. 11 (N=86)
Inst. 12 (N=123)
Inst. 13 (N=82)
Inst. 14 (N=18)
Inst. 15 (N=66)
mg/L CV, % ±2 SD
11 3-26 8-15
26 5-10 21-31
87 3-8 74-100
PT example: Finland, Norway, Canada
pooled or single donation human urine (may be supplemented with albumin and creatinine), liquid, among lab/method comparison
Method N Mean CV ±2SD rangemg/L % mg/L
Finland/Norway 136 19 15.4 14-25
Norway (GPs) 1012 35 12.1 27-44
Canada 28 20 16.5 14-26
Current status of albumin measurement
• Within subject biological variability (CVi) 30-40%
• Limitation: the biological variability data was not uniformly acquired nor analyzed; CVi is difficult to define for disease conditions.
• Many methods have adequate imprecision (CV); some do not (based on PT samples with a uniform molecular species)
• Reliable information on bias among methods is lacking
• Calibration uniformity among methods needs improvement
• The albumin molecular species that are being measured is not well understood
Recommendations: NKDEP/IFCC UA Conf.
• Albumin/Creatinine ratio should be reported
– “mg/mmol” or “mg/g” should be used uniformly in a country or region
• Albumin concentration (mg/L) is difficult to interpret and should not be reported alone
• First morning urine has lower biologic variability than a random collection
• Albumin should be measured on fresh (non-frozen) urine
• “urine albumin” should replace “microalbumin”
Further investigations needed: UA Conf.
• Clarify sample collection and handling requirements for minimum biologic variability
• Clarify the measurand in a urine sample; and the variability of the urine matrix
• Clarify reference intervals and decision points related to risk for kidney damage by age, gender, ethnicity, and concomitant disease
• Develop a reference system– ID/MS measurement procedure at Mayo Clinic (USA)
– Secondary urine reference material from Japan (JSCC and JCCLS)
Outline
• Albumin molecules in urine
• Albumin measurement
• Total protein measurement
Proteins in Urine
• Albumin
• Others• Immunoglobulins
• Bence-Jones
• Tamm-Horsfall
• Lysozyme
• Myoglobin
• Hemoglobin
• Bacterial origin
• Peptides
Quantitative urine protein methods
In order of clinical lab market share in USA:
• Pyrogallol red (dye binding)
• Pyrocatechol violet (dye binding)
• Benzethonium chloride (denaturation/turbidimetry)
• Biuret with precipitation (reference)
• Coomassie blue (dye binding)
Issues with urine protein methods
• Different measurement signal with different proteins
• Different chemical reaction procedures have different response ratios to different proteins
• No uniform calibration standard
Benzethonium chloride Pyrocatechol violet
Pyrogallol red #2Pyrogallol red #1
Dube et al. Clin Biochem 2005;38:479-485.
Normal urines supplemented with: albumin, gamma-globulins, polypeptides, mixed protein
Imai, Clin Chem 1995;32,1986.
Mean total protein of 12 urine samples measured by 7 methods and using 3 standard materials
Methods:SSA – sulfosalicylic acidSSA-SS - sulfosalicylic acid sodium sulfateTCA – trichloroacetic acidBC – benzethonium chlorideCBB – comassie brilliant bluePR-M – Pyrogallol red molybdenumTCA-B - Trichloroacetic acid precipitation biuret
Patients:(3) nephrotic syndrome(1) diabetic nephropathy(1) systemic lupus(1) acute glomerulonephritis(2) multiple meyoloma(4) cancer
Interfering substances with urine protein methods
A non-inclusive list:
• Creatinine
• Aminoglycosides
• Phenothiazines
• Amino acids
• Organic acids
• Peptides
• Other substances
Conclusions
• Urine albumin measurements are less variable than urine total protein measurements
• Clinical trials should use a central lab
• Data from different clinical trials may be difficult to aggregate
Questions
Comments