We would like to assure Selgren (1) that in 2002, the American Diabetes Association could find no published study that fulfilled their requirements of a detection rate of >95% for abnormal samples from patients with microalbuminuria for qualitative (or semiquantitiative) dipstick tests. This was quite independent of measuring albumin by the high-performance liquid chromatography (HPLC) method. We only noted that the American Diabetes Association came to the same conclusion as we did using the HPLC method.

In relation to performance data, the HPLC method has been cleared by the Food and Drug Administration (FDA) (August 2003) as a test to measure albumin. FDA labeling states that “the HPLC method permits a direct measurement of intact albumin, regardless of the reactivity potential of the protein with antibodies. Since immunoassays (and dye binding assays) may not detect all of the intact albumin in urine samples it is expected that the HPLC technology will, depending on the specimen, report greater urinary albumin values when compared to immunochemical urinary albumin test systems and dipstick systems.”

There has been a recent publication in Kidney International (2) demonstrating that the HPLC technique provides earlier detection of kidney disease in humans than the standard albumin assay. The mean lead time for the HPLC assay to detect microalbuminuria versus a radioimmunoassay was 3.9 years for type 1 diabetic patients, with a 95% CI of 2.1–5.6 years. For type 2 diabetic patients, the mean lead time was 2.4 years, with a 95% CI of 1.2–3.5 years.

We emphasize that the HPLC assay measures intact albumin directly, but it will not measure albumin aggregates or albumin-derived fragments. In relation to the Bayer dipstick (microalbustix and CLINITEK microalbumin), we have not seen any published information as to what the dipstick detects (i.e., albumin aggregates, glycated albumin, albumin with lipid or other ligands, or denatured albumin or albumin fragments). All these forms of albumin may exist in urine. All we know is that the Bayer dipsticks give high false-negative rates when compared with an FDA-cleared test that quantitatively detects intact albumin.

1.
Selgren SF: Deficiency in the detection of microalbuminuria by urinary dipstick in diabetic patients (Letter).
Diabetes Care
27
:
2283
,
2004
1.
Comper WD, Osicka TM, Clark M, MacIsaac RJ, Jerums G: Earlier detection of microalbuminuria in diabetic patients using a new urinary albumin assay.
Kidney Int
65
:
1850
–1855,
2004

W.D.C. is employed by and has received grant support from AusAm Biotechnologies. T.M.O. is employed by AusAm Biotechnologies.