| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Letters |
Washington University, Division of Laboratory Medicine, Box 8118, 660 S. Euclid Avenue, St. Louis, MO 63110
a Author for correspondence. Fax 314-362-1461; e-mail mscott{at}labmed.wustl.edu
To the Editor:
Recent studies have clearly demonstrated that plasma homocysteine is an independent risk factor for atherosclerosis (1)(2). Although increased serum homocysteine is genetically determined, increasing dietary folate (3) can significantly lower the values. Indeed, the US Food and Drug Administration has recently increased the daily dietary guidelines for folate intake, and since January 1998 wheat flour-based foods have been supplemented with folate (4)(5). Decreased folate during pregnancy is also a strong risk factor for neurological birth defects (6). Historically, plasma and red blood cell (RBC) folate determinations have been performed to assess folate deficiencies in investigations of macrocytic anemias, and little or no clinical information was ascribed to normal or increased folate values. However, the recent goals of reducing serum homocysteine and maintaining high folate concentrations during pregnancy suggest that examination of serum and RBC folate concentrations with an interest in higher values may become common. Indeed, it has been recommended by some that the lower limit of the reference range for plasma folate be raised by 250% (7). Thus, it is likely that folate concentrations will be increased in the population and that accurate determination of high as well as low folate concentrations will be important.
Historically, there are wide variations in folate analytical
results (8). At our institution, plasma and RBC folate
concentrations are determined using the Abbott AxSYM microparticle
enzyme immunoassay method. We previously had determined that the upper
range of linearity for this method was 14 µg/L by diluting samples
with folate concentrations >25 µg/L with the manufacturer's sample
dilution buffer (unpublished data). The manufacturer claims the method
to be linear to 20 µg/L, but also states in the package inserts
that samples from "some" patients may recover as much as 160%
of the undiluted folate value upon dilution (9). Recently,
we discovered that several samples with folate concentrations of 8–14
µg/L produced values between 12–30 µg/L when they were
"accidentally" repeated using 1:2 dilutions. To investigate this
"serendipitous" finding, we evaluated samples from 526 patients by
performing serum and RBC folate determinations on both undiluted
samples and samples diluted 1:2 using the sample dilution buffer. Fig. 1
clearly demonstrates that a proportionally greater discrepancy
between the undiluted and diluted samples is observed for samples with
undiluted plasma folate values >8 µg/L as assessed by Bland–Altman
plot (10). It is also clear that this greater recovery upon
dilution is not restricted to a subgroup of patient samples, but rather
occurs with all samples having folate values >8–10 µg/L. Indeed,
the average increase in samples with plasma folate values between 8
and 15 µg/L was 134%, whereas it was 168% for samples with folate
values >15 µg/L. Essentially identical findings were observed for
RBC folate samples having a value >10 µg/L on the 1:21 dilution
used to determine RBC folate per the manufacturer's procedure (not
shown).
|
To determine whether the values determined from the undiluted or diluted samples were most accurate, we compared AxSYM values with those obtained by the Bio-Rad Laboratories Quantaphase II 125I-folate radioassay. We determined that the Bio-Rad assay was linear to 20 µg/L (not shown). Twenty samples having values between 3 and 13 µg/L from the Bio-Rad method were compared with values from the AxSYM. For 12 samples, the Bio-Rad values averaged 0.85 µg/L lower than the undiluted AxSYM values. However, eight samples with Bio-Rad values between 11 and 13 µg/L averaged 1.0 µg/L higher than the undiluted AxSYM values, supporting the lack of recovery of the AxSYM method using undiluted samples with values >10 µg/L. On the basis of these findings, we now dilute all plasma and RBC samples 1:2 using Abbott sample dilution buffer when the folate concentration is >8 µg/L on an undiluted sample. Currently, this represents 71% of our samples, making use of the Abbott AxSYM somewhat cumbersome to the laboratory.
Footnotes
Editor's Note: A representative of the manufacturer declined the opportunity to reply
References
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
