CVs for the colorimetric assay and the col- umn method are respectively .... the 95th percentile for. BMG is 0.75 g/L, but discriminant analysis shows that the probability of ... Ann. Med. Interne 127,. 457-460 (1976). J. F. Quaranta'. J. P. Cassuto2.
acid, centrifuge the mixture, mix 2 mL of the supernate with 0.5 mL thiobarbituric acid (50 mmol/L), and incubate at 45 #{176}C for 30 mm. Measure the absorbance. A 10 g/L concentration of Hb A1 corresponds to an absorbance of 0.060 at 488 nm. With fresh hemolysate there is an excellent correlation between the isolated Hb A1 fraction from the column method and data from the colorimetric assay (r = 0.96, n = 50, intra-assay CVs for the colorimetric assay and the column method are respectively 1.5 and 1.7%). The quality control of the colorimetric assay cannot be improved by using 5hydroxymethylfurfural, as has been suggested (7), because the internal standard cannot be correlated to the
of 5-hydroxymethylfurfural the main problem is a break5-hydroxymethylfurfural. Inmakes sense to use a reference such as human Hb (Sigma Co., St. Louis, MO 63178; Human no. H, Type IV, 2X crystallized) or a simple hemolysate from pooled
glycosylated hemoglobin with ion-exchange chromatography. Diabetes 28, 1120-1125 (1979).
In
Th. J. Postmes S. Halders H. Herpers J. P. Sels J. M. Coenegracht Isotope Dept. St. Annadal Hosp., Post bus 1918 Maastricht 6201 BX
The Netherlands
BenceJonesProteinuriaRevisited To the Editor:
The classical heat-precipitation test for detection of Bence Jones proteinuria (1) is now replaced by more specific and sensitive methods such as electrophoresis and various immunodiffusion
blood.
the presence of free monoclonal urinary light chains (FMULC). FMULC in low quantities are not always associated with the diagnosis of malignant monoclonal gammopathy (2).
In our comparative
study with four
graphic method shows an apparent crease in the Hb A1 fraction after sample has been stored for 10 days, creasing linearly by fourfold in
inthe in80
days. In contrast,
was
no sample
instability
found with the colorimetric assay. Again, in experimental animal studies with rabbits we obtained a normal value for Hb A1 with the colorimetric assay (5%), in contrast to the column-chromatographic assay, in which the fresh blood samples showed an Hb A1 that was too low (1.5%). Because of its reliability simplicity, the colorimetric dently is to be preferred
and relative assay evi-
for routine
work. References
1. Trivelli, L. A., Rannay, H. M., and Lai, H. T., Hemoglobin components in patients with diabetes mellitus. N. Engi. I. Med. 284, 353-357 (1971). 2. Rosenthal, M. A., The effect of temperature on the fast hemoglobin test system. Hemoglobin 3, 215 (1979). 3. Jongeneel, J., and van Wissen, M., Optimizing measurement of glycosylated hemoglobins. Ciin. Chem. 25, 642 (1979). Letter. 4. Dunn, P. J., Stuart Soeldner, J., and Wacks, M., Measurement of glycosylated hemoglobins. Lancet ii, 838 (1978). 5. McDonald, J. M., and Davis, J. E., Glycosylated hemoglobins and diabetes mellitus. Hum. Pathol. 10,279-291(1979). 6. Bernstein, R. E., Glycosylated haemoglobin and diabetes. S. Air. Med. J. 55, 821 (1979). 7. Pecoraro, R. E., Graf, R. J., Halter, J. B., et a!., Comparison of a colorimetric assay for 636
CLINICAL CHEMISTRY,
some
cases,
FMULC
may
be
present without any detectable abnormal electrophoretic peak in the serum. In our series of 10 cases, all these patients had malignant diseases (lightchain disease: eight cases; chronic lymphocytic leukemia: one case; nonHodgkin lymphoma one case). A similar observation of one case of myeloblastic leukemia associated with FMULC was elsewhere reported (4). References
amount formed; down of stead, it standard Chemical
hemolysate samples stored at 5#{176}C during 80 days, the column-chromato-
reliable, and selective assay for Bence Jones protein is still lacking.
procedures,
We
which may be used to detect
have
investigated
the
clinical
significance of FMULC in 3850 hospitalized patients. The protein content of urines was measured, and electrophoresis was performed whenever proteinuria was found to exceed 0.5 g/L. Immunoelectrophoresis was performed when an abnormal electrophoretic band was detected in either urine or in serum. All patients’ sera were studied by electrophoresis six myeloma
on cellulose acetate. Fiftycases (including eight cases
of light-chain
72 benign monoclonal gammopathies (BMG), and 10 cases of Waldenstr#{246}m disease were identified.
disease),
FMULC
was observed
in 54%
of the myeloma cases, 20% of the Waldenst#{246}m diseases, and 23% of BMG. If only proteinuria >0.6 g/L were considered (3), the percentage of FMULCpositive myelomas was 45% vs 6% only for BMG. This work confirms the classical oonclusion of the occurrence of FMULC in one myeloma case out of two and one BMG case out of four. It re-emphasizes the importance of the severity of the proteinuria. The discriminative value is 0.6 g/L according to Dammaco and Waldenstr#{246}m (2). However, the 95th percentile for BMG is 0.75 g/L, but discriminant analysis shows that the probability of BMG is only 0.001 for a concentration of 1.2g/L.
Perhaps an even better discrimination may be to use the actual value for Bence Jones protein. All the values described
here refer to total proteinuria, many
cases
may
significantly
which in exceed
Bence Jones protein. However, a simple,
Vol. 27, No. 4. 1981
1. Hence Jones, H., Papers on chemical pathology; prefaced by the Gulstonian lectures, read at the Royal College of Physicians, 1846, Proc. Royal Society, London. Lancet ii, 88-95 (1847). 2. Snapper, I., and Kahn, A., Myelornatosis: Fundamentals and Clinical Features, Karger, Base!, 1971, pp 189-206. 3. Dammaco, F., and Waldenstr#{246}m, J., Hence Jones proteinuria in benign monoclonal gammopathies. Acta Med. Scand. 184, 403-409 (1968). 4. Le Porrier, M., Meyer, 0., Charron, D., and Binet, J. L., Massive and transitory Bence Jones proteinuria in acute myeloblastic leukemia. Ann. Med. Interne 127, 457-460 (1976).
J. F. Quaranta’ J. P. Cassuto2 G. Viot3 P. Audoly2 R. Masseyeff3 1 Lab.
d’H#{233}mato-Cancerologie
(& M. Schneider) Facult#{233} de Med., Chemin de Vallombrose 06034 Nice C#{233}dex, France 2
Ciinique Med. (Pr. P. Audoly)
HOpital
de Cimiez,
Ave. Victoria 06031 Nice C#{233}dex, France 3Lab. d’Immunologie
(Pr. R. Masseyeff) INSERM FRA 12, Facult#{233} de Med., Chemin
de
Vallombrose 06034 Nice
C#{233}dex, France
A Jendrassik-GrolMethod Modified to Eiiminate Hemoglobin
Interferencewith Assayof Total SerumBiiirubin To the Editor: Shull et al. (1) discuss a proposed mechanism to account for interference
by hemoglobin
in determination
of total
bilirubin by the Jendrassik-Grof method. I agree, because results of my experiments are consistent with their
proposal. A sample blank is used to correct the spectral interference due to alkaline
