Apparent loss of urinary albumin during long-term frozen storage: HPLC vs immunonephelometry.

BACKGROUND Urinary albumin detection by immunonephelometry is decreased by approximately 30% in samples that have been frozen at -20 degrees C. An HPLC method for assessment of urinary albumin that detects immunoreactive and immunochemically nonreactive albumin has been introduced as an alternative to immunonephelometry. We investigated whether this technique is affected by sample temperature, particularly freezing. METHODS Urine samples (n = 295) were collected from the general population (Prevention of Renal and Vascular End-Stage Disease Study). Samples were assessed by both immunonephelometry and HPLC when fresh and after storage at -20 degrees C for 4, 8, and 12 months and at -80 degrees C for 12 months. RESULTS With immunonephelometry, storage for 4, 8, and 12 months at -20 degrees C resulted in mean (SD) urine albumin changes of -21% (29%), -28% (29%), and -34% (31) (P <0.001 for trend). Storage at -80 degrees C resulted in a 5% (19%) change after 12 months of storage (not significant). With HPLC, storage for 4, 8, and 12 months at -20 degrees C resulted in urine albumin changes of -33% (28%), -43% (24%), and -55% (21%; P <0.001 vs immunonephelometry). Storage at -80 degrees C resulted in a -29% (29%) change (P <0.001 vs immunonephelometry). CONCLUSION Loss of albumin after freezing urine depends not only on freezing temperature but also on detection method. Detection of albumin by immunonephelometry appears to be significantly less influenced by freezing than detection by HPLC. Storage at -80 degrees C appears to prevent loss when using immunonephelometry, whereas HPLC still shows considerable loss even when urine is frozen at -80 degrees C. We propose that for reliable measurement of urine albumin, fresh samples should be used.