8. Franchini M, Veneri D.

Inherited thrombophilia:
an update. Clin Lab 2005;51:35765.
Giuseppe Lippi
*
Gian Luca Salvagno
Martina Montagnana
Gian Cesare Guidi
Sezione di Chimica
e Microscopia Clinica
Dipartimento di Scienze
Morfologico-Biomediche
Universita` degli Studi di Verona
Verona, Italy
* Address correspondence to this au-
thor at: Istituto di Chimica e Microscopia
Clinica, Dipartimento di Scienze Morfo-
logico-Biomediche, Universita` degli Studi
di Verona, Ospedale Policlinico G.B.
Rossi, Piazzale Scuro, 10, 37134 Verona,
Italy. Fax 39-045-820-1889; e-mail
ulippi@tin.it.
DOI: 10.1373/clinchem.2006.070086
Dilution Test for Differentiating
Falsely High Serum Free
Triiodothyronine Concentrations
To the Editor:
Results of serum assays for free tri-
iodothyronine (FT
3
) can be falsely
increased by cross-reacting drugs
and by autoantibodies (e.g., antithy-
roid hormone autoantibodies, het-
erophilic antibodies). The FT
3
anti-
body included in the Vitros FT3
assay (Ortho-Clinical Diagnostics,
Inc.) cross-reacts with diclofenac (1),
resulting in falsely high measured
FT
3
concentrations. FT
3
and free thy-
roxine (FT
4
) assays that use diiodo-
thyronine (T
2
)-gelatin show falsely
high FT
3
and FT
4
concentrations at-
tributable to the presence of antibod-
ies to T
2
, T
3
, or their conjugates (2).
Concentrations of drugs in blood
are in equilibrium between free and
protein-bound forms. Therefore, ac-
cording to the law of mass action,
serum concentrations of free hor-
mones (FT
3
and FT
4
) do not change
upon dilution of the serum., Thus, a
concentration change in response to
dilution may indicate the presence of
an interfering substance, as reported
by Westerhuis and Venekamp (3),
who found that FT
4
concentrations
decreased when serum containing
anti-T
2
-gelatin and anti-T
3
-gelatin
autoantibodies was diluted 5-fold,
whereas in other sera, FT
4
concentra-
tions showed no change even at 40-
fold dilutions.
We found that measurement of
FT
3
concentrations in diluted sera is
useful for differentiating between the
effects of diclofenac and autoanti-
bodies for the Vitros FT3 assay and
the Vitros FT3II assay (Ortho-Clinical
Diagnostics, Inc.), which does not
show cross-reactivity with diclofenac
(4).
We obtained sera from 2 patients
taking diclofenac and 2 patients [1
reported case (2) and 1 new case]
with antibody against T
2
, T
3
, or their
conjugates. The sera were used for
dilution tests in Vitros FT3 and
Vitros FT3II assays, along with con-
trol sera from 2 healthy individuals
and 2 patients with hyperthyroidism.
The reference interval for the FT
3
value was 4.67.5 pmol/L. We made
serial dilutions (1:1, 1:2, 1:4, and 1:8)
with 0.01 mol/L phosphate-buffered
saline (pH 7.4). The procedures in
this study were in accordance with
the Helsinki Declaration of 1975 and
the amendments of 1996.
Assayed FT
3
concentrations were
5.8 and 5.1 pmol/L in the Vitros FT3
assay and 6.5 and 5.9 pmol/L in the
Vitros FT3II assay for the healthy
individuals, 9.6 and 14.4 pmol/L in
the Vitros FT3 assay and 9.5 and 14.3
pmol/L in the Vitros FT3II assay for
patients with hyperthyroidism,
In the 2 patients taking diclofenac,
FT
3
concentrations were 14.3 and 12.7
pmol/L for the Vitros FT3 assay and
6.6 and 4.6 pmol/L for the Vitros FT3II
assay, indicating that the presence of
diclofenac influenced the Vitros FT3
Fig. 1. Dilution effect of the serum samples on FT
3
value. Sera were diluted with phosphate
buffered saline (pH 7.4). The FT
3
concentration in the diluted serum is plotted as a percentage
of the concentration in nondiluted serum. Tested serum samples are from 2 healthy individuals
(E), 2 patients with hyperthyroidism (, ), 2 patients taking diclofenac (, ), and 2 patients
with antibodies to T
2
, T
3
, or their conjugates (, ). (A), assayed by Vitros FT3. (B), assayed by
Vitros FT3II.
1828 Letters
assay, leading to falsely high values
FT
3
values. FT
3
values in the 2 patients
with antibodies to T
2
, T
3
, or their con-
jugates were 12.3 and 32.6 pmol/L for
the Vitros FT3II assay.
With the Vitros FT3 assay, FT
3
concentrations in 2-fold diluted con-
trol sera were 96% (1.1%) [mean
(SD)] of those in the nondiluted sera,
and in 2-fold diluted sera from 2
patients taking diclofenac, FT
3
con-
centrations were 95.9% and 96.3% of
those in the nondiluted sera (Fig. 1)
With the Vitros FT3II assay, FT
3
con-
centrations in 2-fold diluted control
sera were 97% (0.7%) of those in the
nondiluted sera. For 2-fold diluted
sera from 2 patients taking diclofe-
nac, FT
3
concentrations were 98%
and 97% of those in the nondiluted
sera, indicating that with the Vitros
FT3 assay the dilution effects on FT
3
concentrations did not differ sub-
stantially in sera from patients taking
diclofenac and control sera. This re-
sult suggests that the concentration
of free diclofenac in the bloodstream
does not change with dilution, be-
cause diclofenac is strongly bound to
serum proteins and thus obeys mass
action dilution criteria.
With the Vitros FT3II assay, the
FT
3
concentrations in sera from 2
patients with antibodies to T
2
, T
3
, or
their conjugates decreased 71%
and 55% with dilution. This result
suggests that, in the Vitros FT3II
assay, antibodies to T
2
or its conju-
gate interfered with binding of the
T
2
-confugate. It should be noted that
the converse phenomenon may
occur in the presence of large quan-
tities of weaker binding drugs or
lower, nonsaturating quantities of
autoantibodies.
In conclusion, because the serum
concentration of free drug (diclofe-
nac) in serum is not altered by dilu-
tion, according to the mass action
dilution criteria, the measured FT
3
concentration in the serum remains
falsely high. When the interfering
substance (autoantibody) leading to
falsely high FT
3
concentrations is di-
luted, the interference diminishes
and measured serum FT
3
concentra-
tion decreases. This is the first report
to show that dilution tests can differ-
entiate between drugs and interfer-
ing substances in serum.
This study was supported in part by
a Grant-in-Aid for Encouragement of
Scientists (18925003) from the Japan
Society for the Promotion of Science
(JSPS).
References
1. Kasono K, Hikino H, Fujino S, Takemoto N, Kai T,
Yamaguchi K, et al. Cross-reactive mechanism
for the false elevation of free triiodothyronine in
the patients treated with diclofenac. Endocr J
2001;48:71722.
2. Iwahara K, Tanabe C, Nishiyama K, Ohashi H,
Maekawa M. Falsely high serum free triiodothy-
ronine and free thyroxine concentrations attrib-
utable to anti-diiodothyronine and anti-triiodothy-
ronine antibodies. Clin Chem 2005;51:10712.
3. Westerhuis LW, Venekamp WJ. Falsely high se-
rum free thyroxine concentration measured with
Amerlite-MAB FT4. Clin Chem 1995;41:6334.
4. Iwahara K, Tanabe C, Maekawa M. No interfer-
ence by diclofenac with the new Vitros FT3II
assay reagent. Clin Chem 2004;50:22189.
Kunihiro Iwahara
Chizuko Tanabe
Masato Maekawa*
Department of Laboratory Medicine
Hamamatsu University
School of Medicine
Hamamatsu, Japan
*Address correspondence to this au-
thor at: Department of Laboratory Medi-
cine, Hamamatsu University, School of
Medicine, 1-20-1 Handayama, Ham-
amatsu 431-3192, Japan. Fax 81-53-435-
2794; e-mail mmaekawa@hama-med.ac.
jp.
DOI: 10.1373/clinchem.2006.072272
Dipyrone (Metamizole) Metabolites
Interfere with HPLC Analysis of
Plasma Catecholamines but Not with
the Determination of Urinary
Catecholamines
To the Editor:
Catecholamines and their metabo-
lites are measured for the diagnosis
of pheochromocytoma (1, 2) and
neuroblastoma (3) and for various
other reasons (46). The widely used
HPLC assay, which uses electro-
chemical detection (ECD), combines
high selectivity and sensitivity (7).
The internal standard, 3,4-dihy-
droxybenzylamine (DHBA), is added
before analysis. After pretreatment
with a sample clean-up column, cat-
echolamines are separated and quan-
tified by HPLC-ECD. Commercial
methods (Chromsystems) elute nor-
epinephrine, epinephrine, DHBA,
and dopamine at 8, 9, 1213, and
1719 min, respectively. The next
sample is injected after 25 min.
We sometimes observed an un-
known peak in the plasma catechol-
amine chromatogram that interfered
with the internal standard DHBA
(Fig. 1A). When we repeated the anal-
ysis, the spurious extra peak disap-
peared (Fig. 1B). When we evaluated
the medical histories of these patients,
we found that for a preceding analysis
they had received dipyrone orally or
intravenously within 12 h before blood
collection. Therefore, we hypothesized
that administration of dipyrone and/
or its metabolites may have led to the
interference with DHBA in the subse-
quent analysis.
Dipyrone (metamizole) is widely
used and has effective analgesic, anti-
pyretic, and antispasmodic properties.
After oral or intravenous administra-
tion, dipyrone is rapidly hydrolyzed to
the active moiety 4-methylaminoan-
tipyrine (MAA) (8, 9). MAA is further
metabolized to 4-formylaminoanti-
pyrine (FAA) and 4-aminoantipyrine
(AA), which is acetylated to 4-
acetylaminoantipyrine (AAA). These 4
major metabolites account for 60%of
the administered dose excreted in
urine.
To determine whether dipyrone
and/or its metabolites interfere with
the measurement of plasma cat-
echolamines, we prepared solutions
of dipyrone (1 mg/mL) and MAA,
AA, FAA, and AAA (50 g/mL)
corresponding to 5 times the ex-
pected maximum concentration ob-
tained after a 1-g dose of dipyrone
(9). Dipyrone caused a peak at the
expected retention time of dopamine
(Fig. 1C) and another broad peak in
the next chromatogram at the reten-
tion time of DHBA when an aqueous
injection was started after 25 min
(not shown). The bioactive metabo-
lite MAA also had a retention time of
13 min in the next analysis, and thus
it interfered with the internal stan-
Clinical Chemistry 52, No. 9, 2006 1829