Postcollection Synthesis of Ethyl Glucuronide by
Bacteria in Urine May Cause False Identification of
Alcohol Consumption, Anders Helander,* Ingrid Olsson,
and Helen Dahl (Department of Clinical Neuroscience,
Karolinska Institute and Karolinska University Hospital,
Stockholm, Sweden; * address correspondence to this
author at: Alcohol Laboratory, L7:03, Karolinska Univer-
sity Hospital Solna, SE-171 76 Stockholm, Sweden; fax
46-8-51771532, e-mail firstname.lastname@example.org)
Background: Ethyl glucuronide (EtG) is a minor ethanol
metabolite used as a specific marker to document recent
alcohol consumption; confirm abstinence in treatment
programs, workplaces, and schools; and provide legal
proof of drinking. This study examined if bacterial
pathogens in urine may enable postsampling synthesis
of EtG and ethyl sulfate (EtS) from ethanol, leading to
clinical false-positive results.
Methods: Urine specimens with confirmed growth of
Escherichia coli, Klebsiella pneumoniae, or Enterobacter
cloacae were stored at room temperature in the presence
of ethanol. Ethanol was either added to the samples or
generated by inoculation with the fermenting yeast
species Candida albicans and glucose as substrate. EtG
and EtS were measured by LC-MS.
Results: High concentrations of EtG (24-h range 0.5–17.6
mg/L) were produced during storage in 35% of E. coli-
infected urines containing ethanol. In some specimens
that were initially EtG positive because of recent alcohol
consumption, EtG was also sensitive to degradation by
bacterial hydrolysis. In contrast, EtS was completely
stable under these conditions.
Conclusions: The presence of EtG in urine is not a
unique indicator of recent drinking, but might originate
from postcollection synthesis if specimens are infected
with E. coli and contain ethanol. Given the associated
risks for false identification of alcohol consumption and
false-negative EtG results due to bacterial degradation,
we recommend that measurement of EtG be combined
with EtS, or in the future possibly replaced by EtS.
© 2007 American Association for Clinical Chemistry
Early recognition of problem drinking or relapse is im-
portant to ensure adequate alcohol treatment strategies
(1). This goal has been hampered by a lack of sufficiently
sensitive and specific diagnostic methods. The reliability
of self-reporting is limited by denial and underreporting
(2). The time frame for identifying alcohol use by ethanol
testing is usually limited to ?12 h, because of rapid
metabolism and excretion (3). Research has therefore
focused on developing alcohol biomarkers with a longer
detection window (4).
A new laboratory marker for detecting recent alcohol
consumption is ethyl glucuronide (EtG) (5). EtG and ethyl
sulfate (EtS) (6) are minor ethanol metabolites formed by
uridine diphosphate–glucuronosyltransferase (UGT) and
sulfotransferase (SULT), respectively, and excreted in
urine for a longer time than ethanol (7–10). Positive EtG
and/or EtS test results thus provide a strong indication
that the person has recently consumed alcohol, even when
ethanol is no longer detectable (9). LC-MS methods are
available for EtG and EtS detection (6, 10), as is an
enzyme immunoassay for EtG (DRI®EtG, Microgenics).
EtG has been recommended for forensic application
(11–13) and is used for documentation of abstinence in
treatment programs, for alcohol testing in the workplace
and schools, and as legal proof of drinking (known as “the
80-h alcohol test”). However, the high diagnostic sensitiv-
ity of the EtG test has also produced adverse publicity
(14), because unintentional ethanol intake from ethanol-
based mouthwash (15) and hand sanitizers (16) may also
generate positive results. The United States Substance
Abuse and Mental Health Services Administration re-
cently warned against using a positive EtG as primary or
sole evidence of drinking for disciplinary and legal action
Bacterial contamination of urine may cause false-nega-
tive EtG test results (18). Many strains of Escherichia coli,
the main source of urinary tract infections, contain the
enzyme ?-glucuronidase, which hydrolyzes EtG. Given
that UGT and SULT activity also occur with some bacteria
(19, 20), we examined whether human pathogens may
enable postcollection synthesis of EtG and EtS from
ethanol in urine.
Fresh human urine specimens (anonymous surplus
volumes) with confirmed growth of common pathogenic
bacteria (E. coli, n ? 36; Klebsiella pneumoniae, n ? 6;
Enterobacter cloacae, n ? 6), as identified by culture on
standard solid media, were used (study approved by the
local ethics committee). The samples had been submitted
for routine diagnostic testing in the Department of Clini-
cal Microbiology, Karolinska University Hospital, and
were stored refrigerated until use.
In the 1st experiment we added ethanol (final concen-
tration 1.0 g/L) to urine samples and split them into tubes
that were capped and stored at 4 °C and 22 °C. The same
samples without addition of ethanol, or supplemented
with ethanol and 10 g/L sodium fluoride as preservative,
and uninfected urines served as controls. In the 2nd
experiment ethanol was generated in the urine samples by
inoculation with the fermenting yeast species Candida
albicans (1 000 000 colony-forming units/L) and 20 g/L
glucose as substrate. At the start of the experiment, and
after different storage times at 4 °C and 22 °C, aliquots
were stored at ?20 °C before analysis of EtG, EtS, and
EtG and EtS were quantified by an LC-MS method
(6, 9, 10). Analysis was performed in the negative-ion
mode using selected ion monitoring of the deprotonated
ions at m/z 125 for EtS and m/z 130 for EtS-D5, and at m/z
221 and m/z 226 for EtG and EtG-D5. We purchased EtS
from TCI and EtG and EtG-D5 from Medichem Diagnos-
tics. EtS-D5 was synthesized (9). The previously deter-
mined detection limit was 0.1 mg/L; the routine clinical
cutoff for EtG used in our laboratory is 0.5 mg/L. All
positive EtG results by LC-MS were confirmed by LC-
Clinical Chemistry 53, No. 10, 2007
tandem MS (Perkin–Elmer 200 LC and Sciex API 2000 MS)
by the presence of the correct relative abundance of the
major product ions of EtG (m/z 75, 85, and 113). No
interference by ion suppression was noted.
The ethanol concentration was determined enzymati-
cally using alcohol dehydrogenase on a Hitachi 917
Of the 36 urine specimens infected by E. coli, 10 were
positive for EtG (range 2.6–135.9 mg/L, mean 25.7 mg/L,
median 10.4 mg/L) and EtS (range 1.3–20.0 mg/L, mean
5.0 mg/L, median 3.5 mg/L) at the start of the experi-
ment, indicating that these patients had recently con-
sumed alcohol. After these 10 samples were stored for 5
days at 22 °C, EtG was no longer detectable in 5 (50%),
whereas the EtS concentrations remained unchanged. A
disappearance of EtG, but not of EtS, was also observed
after the samples had been supplemented with ethanol
(Table 1). In 3 samples that initially contained 11.7–46.6
mg/L EtG, the concentrations were below the routine
clinical cutoff (?0.5 mg/L) after 24-h storage at 22 °C.
These samples also showed a gradual disappearance of
EtG at 4 °C, albeit at a much slower rate, whereas sodium
fluoride was effective in preventing EtG degradation both
at 4 °C and 22 °C (data not shown).
In 9 (35%) of the 26 urine specimens with confirmed
growth of E. coli that were initially negative for EtG and
EtS, formation of EtG but not of EtS was observed with
time at 22 °C after addition of 1 g/L ethanol. In 7 samples
(Fig. 1A), EtG concentrations above the clinical cutoff
were observed after 24-h storage (range 0.5–17.6 mg/L,
mean 5.2 mg/L, median 3.3 mg/L), and after 5 days the
concentrations ranged from 0.3 to 35.2 mg/L (mean 8.9
mg/L, median 2.4 mg/L, n ? 9). Slow formation of EtG
was also observed in 3 samples in the presence of added
sodium fluoride and in 2 samples stored at 4 °C. After
addition of ethanol to 1 urine specimen that initially
contained 8.9 mg/L EtG and 2.0 mg/L EtS, the EtG
concentration first increased to 17.6 mg/L after 24-h
storage but then decreased to 5.6 mg/L after 48 h and then
to ?0.5 mg/L after 5 days (Fig. 1B). The corresponding
EtS concentrations were stable at all times.
Table 1. Stability of the minor ethanol metabolites EtG and
EtS during storage of infected urine specimens.
Test conditions Uropathogen identified
6 Infected urine specimens
incubated with 1.0
g/L ethanol for 5
positive for EtG and
EtG unchanged after 5
EtG negative or
decreased after 5
EtS unchanged after 5
EtS negative or
decreased after 5
negative for EtG and
EtG negative after 5
EtG positive after 5
EtS negative after 5
EtS positive after 5
aEthanol was added to fresh urine specimens with confirmed growth of E. coli,
K. pneumoniae, or E. cloacae and stored in sealed plastic vials without
preservative at 22 °C for 5 days.
bSamples initially positive for EtG and EtS indicated that these patients had
recently consumed alcohol. EtG and EtS were measured by LC-MS and the
detection limit was approximately 0.1 mg/L for both compounds.
cAll positive LC-MS results were confirmed positive by liquid chromatography-
tandem mass spectrometry.
Fig. 1. Formation of the ethanol metabolite EtG in E. coli infected urine
samples after addition of ethanol.
(A), urine specimens with confirmed growth of E. coli were supplemented with 1.0
g/L ethanol and stored at 22 °C. Individual EtG results for 7 urine samples are
indicated by different symbols (the data for 3 samples that produced low
concentrations of EtG partly overlap). (B), results for 1 urine specimen with
confirmed growth of E. coli that showed both synthesis and degradation of EtG.
The sample was initially positive for EtG (?) and EtS (F) and showed variable EtG
but unchanged EtS concentrations with time after addition of ethanol and
storage at 22 °C.
After 7 urine specimens containing E. coli and 5 unin-
fected control urines were supplemented with C. albicans
and glucose to generate ethanol (all samples were initially
negative for ethanol), the ethanol concentrations after
7-day storage at 22 °C ranged from 0.73 to 1.47 g/L
(median 1.17 g/L). Formation of EtG (range 1.8–71.4
mg/L) was observed in 3 specimens containing E. coli, but
in none of the uninfected controls. No formation of EtS
was detected in these experiments.
No disappearance or formation of EtG or EtS was
observed in the 12 urine specimens with confirmed
growth of K. pneumoniae or E. cloacae after addition of
ethanol or C. albicans and glucose and storage at 22 °C for
5 days (Table 1).
EtG has been considered specific for alcohol consump-
tion and detectable only after in vivo ethanol metabolism,
and hence EtG testing is used as a basis for disciplinary
and legal action (17) and in forensic autopsy cases (13). A
recent debate relates to the excellent analytical sensitivity
of this test that, in combination with a low clinical cutoff
concentration, may cause positive results attributable to
unintentional ethanol exposure (14–16). To the best of our
knowledge, no true false-positive EtG result has been
reported without such exposure. Nonetheless, the present
study demonstrated that EtG could be formed in a bio-
logical specimen after collection, if the specimen is in-
fected with E. coli and ethanol is present or produced
during storage. In our tested samples the formation of EtG
was rapid and was not always prevented by addition of
sodium fluoride or storage at refrigerator temperature.
Bacterial and fungal infections are common in clinical
practice, with E. coli being the primary pathogen respon-
sible for urinary tract infections. Ethanol may be formed
in unpreserved biological specimens because of microbial
contamination and fermentation, and this risk is espe-
cially high in diabetic patients as a result of glycosuria.
Accordingly, considering the potential serious disciplin-
ary and legal consequences if an individual is falsely
accused of alcohol consumption on the basis of an incor-
rect EtG result, caution is advised when interpreting EtG
test results, and the risk for postcollection ethanol forma-
tion must be considered.
The results of our study also confirm previous obser-
vations that EtG is sensitive to bacterial hydrolysis, but
EtS is not (18). Accordingly, in situations in which EtG-
positive urine is infected from the start, or becomes
contaminated during handling, there is a risk for false-
negative results and alcohol use will remain undetected.
The lack of EtS formation or degradation detected
under the test conditions and the similar detection win-
dows and sensitivities for recent alcohol consumption
observed for the unique ethanol metabolites EtG and EtS
(9) indicate that EtS testing should accompany, be used to
verify, or in the future possibly replace EtG testing. The
results further indicate that EtS is a more suitable test than
EtG to distinguish antemortem ingestion of ethanol from
postmortem synthesis in forensic toxicological analysis
(13). Mass spectrometric methods for EtG can easily be
modified to also quantify EtS (6, 9). If the analysis initially
focuses solely on EtG, EtS may be introduced as a
verification assay. However, a false-negative EtG screen-
ing result will usually not be followed up with confirma-
tory analysis, anddrinking
will thereby remain
Grants/funding support: The present work was funded by
the Karolinska Institute and the Stockholm County Council,
Financial disclosures: None declared.
1. Babor TF, Aguirre-Molina M, Marlatt GA, Clayton R. Managing alcohol
problems and risky drinking. Am J Health Promot 1999;14:98–103.
2. Fuller RK, Lee KK, Gordis E. Validity of self-report in alcoholism research:
results of a Veterans Administration cooperative study. Alcohol Clin Exp Res
3. Helander A, Beck O, Jones AW. Laboratory testing for recent alcohol
consumption: comparison of ethanol, methanol, and 5-hydroxytryptophol.
Clin Chem 1996;42:618–24.
4. Helander A. Biological markers in alcoholism. J Neural Transm Suppl
5. Seidl S, Wurst FM, Alt A. Ethyl glucuronide: a biological marker for recent
alcohol consumption. Addict Biol 2001;6:205–12.
6. Helander A, Beck O. Mass spectrometric identification of ethyl sulfate as an
ethanol metabolite in humans. Clin Chem 2004;50:936–7.
7. Schmitt G, Aderjan R, Keller T, Wu M. Ethyl glucuronide: an unusual ethanol
metabolite in humans. Synthesis, analytical data, and determination in
serum and urine. J Anal Toxicol 1995;19:91–4.
8. Dahl H, Stephanson N, Beck O, Helander A. Comparison of urinary excretion
characteristics of ethanol and ethyl glucuronide. J Anal Toxicol 2002;26:
9. Helander A, Beck O. Ethyl sulfate: a metabolite of ethanol in humans and
a potential biomarker of acute alcohol intake. J Anal Toxicol 2005;29:
10. Stephanson N, Dahl H, Helander A, Beck O. Direct quantification of ethyl
glucuronide in clinical urine samples by liquid chromatography-mass spec-
trometry. Ther Drug Monit 2002;24:645–51.
11. Wurst FM, Skipper GE, Weinmann W. Ethyl glucuronide: the direct ethanol
metabolite on the threshold from science to routine use. Addiction 2003;
12. Skipper GE, Weinmann W, Thierauf A, Schaefer P, Wiesbeck G, Allen JP, et
al. Ethyl glucuronide: a biomarker to identify alcohol use by health profes-
sionals recovering from substance use disorders. Alcohol Alcohol 2004;39:
13. Kugelberg FC, Jones AW. Interpreting results of ethanol analysis in postmor-
tem specimens: a review of the literature. Forensic Sci Int 2007;165:10–
14. Helliker K. A test for alcohol and its flaws: a new screen detects Sunday’s
gin in Monday’s urine but it may be ensnaring some innocent people too.
Beer or hand sanitizer? Wall St J, August 12, 2006:A1.
15. Costantino A, Digregorio EJ, Korn W, Spayd S, Rieders F. The effect of the
use of mouthwash on ethylglucuronide concentrations in urine. J Anal
16. Rohrig TP, Huber C, Goodson L, Ross W. Detection of ethylglucuronide in
urine following the application of Germ-X. J Anal Toxicol 2006;30:703–4.
17. Center for Substance Abuse Treatment. The role of biomarkers in the
treatment of alcohol use disorders. Substance Abuse Treatment Advisory
18. Helander A, Dahl H. Urinary tract infection: a risk factor for false-negative
urinary ethyl glucuronide but not ethyl sulfate in the detection of recent
alcohol consumption. Clin Chem 2005;51:1728–30.
19. Bock KW. Vertebrate UDP-glucuronosyltransferases: functional and evolu-
tionary aspects. Biochem Pharmacol 2003;66:691–6.
20. Chen WT, Liu MC, Yang YS. Fluorometric assay for alcohol sulfotransferase.
Anal Biochem 2005;339:54–60.
Previously published online at DOI: 10.1373/clinchem.2007.089482
Clinical Chemistry 53, No. 10, 2007
Stability of Soluble Adhesion Molecules, Selectins,
and C-Reactive Protein at
Clinical Studies, Janine Hartweg,1,2*Michael Gunter,1Rafael
Perera,2Andrew Farmer,1,2Carole Cull,1†Casper Schalkwijk,3‡
Astrid Kok,3Harry Twaalfhoven,3Rury Holman,1and Andrew
Neil1,2(1Diabetes Trials Unit, Oxford Centre for Diabetes,
Endocrinology and Medicine, University of Oxford,
United Kingdom;2Division of Public Health and Primary
Health Care, University of Oxford, United Kingdom;
3Clinical Chemistry, Institute for Cardiovascular Re-
search VU University Medical Centre, Amsterdam, The
Netherlands; † Dr. Carole Cull died in June 2007; ‡ current
address: Department of Internal Medicine, University
Hospital Maastricht, Maastricht, The Netherlands; * ad-
dress correspondence to this author at: Diabetes Trials
Unit, Oxford Centre for Diabetes, Endocrinology and
United Kingdom; fax 44 (0)1865 857240, e-mail janine.
Background: We assessed the impact of sample storage
conditions on soluble vascular cell adhesion molecules
(sVCAM), soluble intracellular adhesion molecules
(sICAM-1), soluble (s)E-selectin, C-reactive protein
(CRP), and sP-selectin.
Methods: Markers were measured by ELISA in venous
blood from 10 healthy volunteers on aliquots stored as
plasma or whole blood at 4, 21, or 30 °C for 1–5 days and
after 1–5 freeze-thaw cycles. We compared results on
these samples to results for samples processed immedi-
ately and stored at ?80 °C. Statistical models assessed
Results: Using an upper limit of 10% variation from
baseline with P >0.05, we found that stability duration
in plasma was 5 days for sVCAM-1 and sICAM-1 and at
least 2 days for sE-selectin at 4, 21, and 30 °C and 5 days
for CRP at 4 and 21 °C and 1 day at 30 °C. Stability
duration in whole blood was 5 days for sVCAM-1 and
sICAM-1 and at least 2 days for sE-selectin at 4, 21, and
30 °C and 5 days for CRP at 4 and 21 °C and 2 days at
30 °C. sP-selectin was not stable in plasma or whole
blood. sICAM-1, sVCAM-1, CRP, and sE-selectin were
stable after 5 freeze-thaw cycles.
Conclusions: sVCAM-1, sICAM-1, and CRP are stable
in plasma or whole blood at 4 and 21 °C for at least 3
days and sE-selectin for 2 days. sP-selectin is not stable
and therefore requires immediate assay.
© 2007 American Association for Clinical Chemistry
effects of postprocessing
Multicenter studies that measure the cardiovascular risk
(sICAM-1), soluble vascular cell adhesion molecules-1
(sVCAM-1), soluble (s)E-selectin, sP-selectin, and C-reac-
tive protein (CRP) in serum or plasma require validated
procedures for processing, storage, and transport of sam-
ples. Previous studies have not systematically examined
effects on these markers of different storage temperatures,
duration of storage, freeze-thaw cycles, or delays in
separating plasma from whole blood (1–5). Venous blood
samples were obtained from 10 nonsmoking healthy staff
members (7 male, 3 female, age 24–58 years) who had
given informed consent. Clinical research nurses collected
the samples into 3 7-mL K3 EDTA glass tubes and 12 2-mL
K3 EDTA glass tubes (Becton Dickinson) per person.
Samples were randomized into 3 groups. Group 1 sam-
ples [six 0.5-mL samples from 5 volunteers (4 males)
(see Fig. 1 in the Data Supplement that accompanies
this article at http://www.clinchem.org/content/vol53/
issue10)] were centrifuged immediately, stored at ?80 °C,
and then subjected to 0–5 room temperature freeze-thaw
cycles. Group 2 samples [1 2.5-mL samples from each
volunteer (Supplementary Data Fig. 2)] were centrifuged
immediately, and the plasma was then divided into 13
aliquots. One aliquot was frozen immediately, and the
others were maintained at 4, 21, or 30 °C for 1, 2, 3, or 5
days and then frozen at ?80 °C until assayed. Group 3
samples [1 3-mL sample from each volunteer (Supple-
mentary Data Fig. 2)] were split into 13 aliquots, 1 aliquot
was centrifuged immediately and the plasma stored at
?80 °C, and the remaining 12 aliquots were stored as
whole blood under the same conditions as group 2 before
centrifugation and freezing of the plasma at ?80 °C.
Frozen aliquots were packed in dry ice and sent to the VU
University Medical Centre, The Netherlands, where sam-
ples were thawed, inverted, and vortex mixed before
All aliquots were assayed in duplicate at room temper-
ature in batches of 39 samples in up to 2 runs per
sampling method for each analyte, including the freeze-
thaw aliquots, using semiautomated ELISA methods for
sVCAM-1, sICAM-1, and high-sensitivity CRP (all ob-
tained from Diaclone) on the CODA automated EIA
reader (Bio-Rad) and manual ELISA methods for sE-
selectin (Diaclone) and sP-selectin (R&D Systems) (6).
Baseline values used for each analyte were those ob-
tained from assay of plasma from samples that had been
separated immediately. The mean percentage change
from baseline was calculated for each time point, temper-
ature, and freeze-thaw cycle using the Statistical Package
for the Social Sciences (v12.2, SPSS). Analyte values that
differed by ?10% from baseline and were statistically
significant (P ?0.05) were considered to indicate unac-
ceptable stability for given storage/handling conditions
(7). We used generalized estimating equations to assess
correlations in the data and compared the different stor-
age time and temperature values to baseline values for
each sample (8). P values reported are those from gener-
alized estimating equation models obtained using the
statistical package STATA (Intercooled STATA 8.2 for
Mean (range) baseline values for sVCAM-1 were 873
(715–1283) ?g/L, sICAM-1 491 (350–864) ?g/L, sE-selec-
tin 41 (31–75) ?g/L, high-sensitivity CRP 4 (0.17–12.15)
mg/L, and sP-selectin 29 (11.5–40.1) ?g/L. These values
were all within published reference intervals (9–12) after
removal of probable outliers (values ?4 SD from the
mean) (13). Interassay CVs ranged from 1.3% to 2.4% at
660 to 703 ?g/L for sVCAM-1, 3.6% to 5.6% at 348 to 485
?g/L for sICAM-1, 9.9% to 10.0% at 14.2 to 18.3 ?g/L for
sE-selectin, 3.0% to 6.1% at 0.78 to 1.59 mg/L for CRP, and
9.0% to 9.4% at 30.2 and 38.2 ?g/L for sP-selectin.
Intraassay CVs were 4.4% for sVCAM-1, 4.0% for
sICAM-1, 4.0% for sE-selectin, 3.9% for CRP, and 2.7% for
sP-selectin. VCAM-1 and sICAM-1 were stable up to 5
days under all storage conditions in plasma and whole
blood, and after 5 freeze-thaw cycles (Table 1), although
whole blood gave more reproducible results than plasma
for sICAM-1. sE-selectin was stable for up to 2 days at
4 °C in plasma and whole blood and for 5 freeze-thaw
cycles. CRP was stable for 5 days at 4 and 21 °C in plasma
and whole blood and for 5 freeze-thaw cycles. sP-selectin
was unstable under all storage conditions and thus re-
quires immediate processing (Fig. 1 and Table 1) (also see
Supplemental Data Table 1 and Supplemental Data Table
2 for all results).
This study is the 1st to assess the effects of sample storage
for up to 5 days at different temperatures and exposure to
repeat freeze-thaw cycles on stability of soluble adhesion
molecule and selectin measurements in whole blood and
plasma. Previous studies of CRP stability (1, 2, 5, 8, 14–16)
did not consider repeated freeze-thaw cycles or higher
storage temperatures beyond 3 days in plasma vs whole
blood, and investigated only the effects of storage beyond 5
days at 4 or 21 °C. Our sample size was too small to assess
concentration-dependent effects, which should be assessed
in a similar manner in future studies.
Assays were conducted at an ambient temperature of
21 °C in single batches by 2 laboratory technicians. To
improve precision and run-to-run variability, 1 technician
used semiautomated analysis methods on 3 of the markers
Table 1. Number days for which analyte values did not differ significantly (P >0.05) from baseline, with <10% variation
after storage as plasma or whole blood at 4 °C, 21 °C, or 30 °C or after a number of freeze-thaw cycles.
Plasma (days)4 °C5
Whole blood (days)4 °C5
Freeze-thaw cycles (n)
Fig. 1. (A), mean (95% CI) percentage change from baseline for each analyte after 1–5 freeze-thaw cycles.
Values that differ significantly from baseline (P ?0.05) are denoted by an asterisk. (B), mean (95% CI) percentage day-to-day change for days 0–1, 1–2, 2–3, and 3–5
for each analyte, calculated using robust variance estimates and adjusting for plasma or whole blood samples showing variation after storage at prespecified
temperatures (E ? 4 °C plasma; F ? 4 °C whole blood; ‚ ? 21 °C plasma; Œ ? 21 °C whole blood; v ? 30 °C plasma; f ? 30 °C whole blood).
Clinical Chemistry 53, No. 10, 2007
(sVCAM, sICAM, CRP), and manual methods were used on
the other assayed selectins. A few measurements do not fit
any discernable pattern, notably plasma sVCAM-1 on day 3
at 4 and 21 °C; plasma sE-selectin on day 1 at 4 °C and day
3 at all temperatures, whole blood sE-selectin at 4 °C on day
1 and after freeze-thaw cycle 4; and plasma CRP on day 1 at
4 °C and after freeze-thaw cycle 3. These aberrant values
were likely analytic artifacts, because removal of values ?4
SDs from the mean or adjustment for run-to-run variability
did not change the results qualitatively. The large CRP
difference observed after freeze-thaw cycle 3 but not after
cycles 4 or 5 suggests assay error or possibly CRP release
from LDL and complement factors. It has been demon-
strated that a portion of systemic CRP is bound to choles-
terol in modified LDL particles (17) and to different com-
plement factors (18). Interaction of CRP with cholesterol and
complement factors might hamper the detection of CRP in
the assay used. No data were available for sE-selectin, but it
is possible that some sE-selectin is bound to different circu-
lating leukocyte types or microvesicles derived from endo-
thelial cells. Freeze-thaw cycles might release these bound
forms of sE-selectin, leading to an increase in detectable
Possible effects of residual platelets in the plasma on
marker values obtained are not addressed in our study.
sVCAM-1, sICAM-1, and sE-selectin, however, are de-
rived mainly from endothelial cells and not from platelets.
Although the presence of platelets in EDTA plasma might
contribute to the absolute amounts of these markers, it
does not influence their stability. In contrast, sP-selectin is
mainly derived from platelets, a characteristic that may
explain the variance observed in sP-selectin at different
time and temperature points.
Our results show that sP-selectin is unstable under all
storage conditions and requires immediate assay. The
other cardiovascular risk markers evaluated were stable
when stored in whole blood samples for several days at
room temperature. sVCAM-1, sICAM-1, and CRP were
stable at 4 and 21 °C in plasma or whole blood for 5 days
and sE-selectin for 2 days. These findings have important
implications for clinical studies measuring these markers,
reducing the need for immediate transfer of samples to a
laboratory for processing and analysis.
Grant/funding support: This study was funded jointly by the
University of Oxford’s Division of Public Health and Primary
Health Care and Diabetes Trials Unit.
Financial disclosures: None declared.
Acknowledgments: We thank Dr. Brian Shine for assistance
with sensitivity analyses of run-to-run variability and quality
1. Aziz N, Fahey JL, Detels R, Butch AW. Analytical performance of a highly
sensitive C-reactive protein-based immunoassay and the effects of labora-
tory variables on levels of protein in blood. Clin Diagn Lab Immunol
2. Laurent P, Marchand B, Bienvenu J, Marichy J. Rapid enzyme immunoassay
for quantification of C-reactive protein (CRP). Clin Biochem 1985;18:272–5.
3. Lozano M, Estebanell E, Cid J, Diaz-Ricart M, Mazzara R, Ordinas A, et al.
Platelet concentrates prepared and stored under currently optimal condi-
tions: minor impact on platelet adhesive and cohesive functions after
storage. Transfusion 1999;39:951–9.
4. Mondoro TH, Shafer BC, Vostal JG. Restoration of in vitro responses in
platelets stored in plasma. Am J Clin Pathol 1999;111:693–9.
5. Pai JK, Curhan GC, Cannuscio CC, Rifai N, Ridker PM, Rimm EB. Stability of
novel plasma markers associated with cardiovascular disease: processing
within 36 hours of specimen collection. Clin Chem 2002;48:1781–4.
6. De Jager J, Kooy A, Lehert P, Bets D, Wulffele MG, Teerlink T, et al. Effects
of short-term treatment with metformin on markers of endothelial function
and inflammatory activity in type 2 diabetes mellitus: a randomized, placebo-
controlled trial. J Intern Med 2005;257:100–9.
7. Roberts WL, Schwarz EL, Ayanian S, Rifai N. Performance characteristics of
a point of care C-reactive protein assay. Clin Chim Acta 2001;314:255–9.
8. Zeger SL, Liang KY. Longitudinal data analysis for discrete and continuous
outcomes. Biometrics 1986;42:121–30.
9. Macy EM, Hayes TE, Tracy RP. Variability in the measurement of C-reactive
protein in healthy subjects: implications for reference intervals and epide-
miological applications. Clin Chem 1997;43:52–8.
10. Costongs GM, Janson PC, Brombacher PJ. Effects of biological and analyt-
ical variations on the appropriate use of “reference intervals” in clinical
chemistry: proposal of a scheme for longitudinal assessment of laboratory
values. J Clin Chem Clin Biochem 1984;22:613–21.
11. Gearing AJH, Newman W. Circulating adhesion molecules in disease.
Immunol Today 1993;14:506–12.
12. Ridker PM, Rifai N, Cook NR, Bradwin G, Buring JE. Non-HDL cholesterol,
apolipoproteins A-I and B100, standard lipid measures, lipid ratios, and CRP as
risk factors for cardiovascular disease in women. JAMA 2005;294:326–33.
13. Cull CA, Manley SE, Stratton IM, Neil HA, Ross IS, Holman RR, et al.
Approach to maintaining comparability of biochemical data during long-term
clinical trials. Clin Chem 1997;43:1913–8.
14. Kayaba K, Ishikawa S, Gotoh T, Nago N, Kajii E, Nakamura Y, et al. Five-year
intra-individual variability in C-reactive protein levels in a Japanese popula-
tion-based study: the Jichi Medical School Cohort Study at Yamato, 1993–
1998. Jpn Circ J 2000;64:303–8.
15. Sarikaputi M, Morimatsu M, Yamamoto S, Syuto B, Saito M, Naiki M. Latex
agglutination test: a simple, rapid and practical method for bovine serum
CRP determination. Jpn J Vet Res 1992;40:1–12.
16. Link LB, Jacobson JS. Effect of delayed processing on high-sensitivity
C-reactive protein. Clin Chem 2006;52:787–8.
17. Taskinen S, Hyvonen M, Kovanen PT, Meri S, Pentikainen MO. C-reactive
protein binds to the 3beta-OH group of cholesterol in LDL particles. Biochem
Biophys Res Commun 2005;329:1208–16.
18. Wolbink GJ, Brouwer MC, Buysmann S, ten Berge IJ, Hack CE. CRP-mediated
activation of complement in vivo: assessment by measuring circulating
complement-C-reactive protein complexes. J Immunol 1996;157:473–9.
Previously published online at DOI: 10.1373/clinchem.2006.076380
Genomic Profiling of Circulating Plasma RNA for the
Analysis of Cancer, Manuel Collado,1Vanesa Garcia,2Jose
Miguel Garcia,2Isabel Alonso,2Luis Lombardia,1Ramon Diaz-
Uriarte,1Luis A. Lo ´pez Ferna ´ndez,3Angel Zaballos,4Fe ´lix
Bonilla,2and Manuel Serrano1*(1Spanish National Cancer
Research Centre (CNIO), Madrid, Spain;2Department of
Oncology, Hospital Universitario Puerta de Hierro, Ma-
drid, Spain;3Department of Pharmacogenetics and Phar-
macogenomics, Hospital Universitario Gregorio Mara-
n ˜o ´n, Madrid, Spain;4National Centre of Biotechnology
(CNB-CSIC), Campus Universidad Auto ´noma, Madrid,
Spain; * address correspondence to this author at: Spanish
National Cancer Research Centre (CNIO), 3 Melchor
Ferna ´ndez Almagro St., 28029 Madrid, Spain; fax 34-91-
732-8028, e-mail email@example.com)
Background: The blood of cancer patients is known to
contain fragments of RNA released from the tumor. The
application of genomic profiling techniques to plasma
RNA may allow the unbiased selection of cancer mark-
ers in the blood, but the informative value of genomic
profiling of plasma RNA is currently unknown.
Methods: We used cDNA microarray hybridization to
perform genomic profiling of plasma RNA from colo-
rectal cancer (CRC) patients and from healthy donors.
From a list of 40 genes differentially upregulated in
cancer patients, we randomly selected 4 genes for fur-
ther characterization. These 4 markers were analyzed by
quantitative reverse-transcription PCR in a wide set of
samples including paired samples from the same CRC
patients before and after surgical resection of the tumor.
UBE2D3, and KIAA0101—were confirmed by PCR to be
significantly increased in cancer compared to healthy
donors. Importantly, 2 of the markers, EPAS1 and
UBE2D3, showed a significant decrease after surgery,
returning to the levels of healthy donors. Finally, super-
vised class prediction using these 3 markers correctly
(77%) assigned presurgery samples to the CRC group
and assigned postsurgery samples from the same pa-
tients to the healthy group.
Conclusions: Our findings demonstrate the usefulness
of gene expression profiling of circulating plasma RNA
to find cancer markers of potential clinical value.
© 2007 American Association for Clinical Chemistry
The blood of cancer patients contains higher concentra-
tions of DNA than does the blood of healthy individuals
(1). The development of PCR amplification techniques
has allowed the analysis of this circulating DNA, and a
large body of evidence has demonstrated that the plasma
DNA from cancer patients presents features of the cancer
DNA, suggesting that it derives from tumor cells (2–5).
More recently, several groups have reported the extrac-
tion of RNA from the plasma of cancer patients and its
subsequent analysis by reverse transcription (RT)-PCR
(6–10). The potential use of plasma RNA for the analysis
of cancer is highly attractive for several reasons: it re-
quires a minimally invasive method (collection of a small
amount of blood); it can be obtained at any time and in a
repetitive fashion, allowing the analysis of disease pro-
gression and treatment response; and its simplicity is
amenable for use in asymptomatic populations at risk.
The analysis of plasma RNA has been restricted to a few
markers assumed to be abundant and specifically associ-
ated with particular cancer types, for example, mamma-
globin for breast cancer and tyrosinase for melanoma
(7, 10). Further progress toward the clinical use of plasma
RNA requires the unbiased identification of markers.
Genome-wide profiling of plasma RNA is an obvious
approach but has technical drawbacks that could prevent
its application, such as the low abundance and lack of
integrity of plasma RNA (11).
For this reason we evaluated the feasibility of a genomic
approach to studying plasma RNA. We measured by
cDNA microarray hybridization the relative abundance of
the different RNA species in the plasma of colorectal
cancer (CRC) patients (n ? 12) and healthy donors (n ? 8).
All the patients and healthy donors along this study gave
their informed consent following the rules of the Research
Ethics Board of Hospital Universitario Puerta de Hierro.
Each sample was competitively hybridized against a
common reference formed by a pool of blood samples
from 26 healthy donors different from those hybridized as
individual healthy samples. Differential gene expression
analysis between CRC and healthy donors identified a
total of 87 genes, including 40 that were differentially
upregulated in the cancer group (see Supplementary Fig.
1A and Supplementary Table 1 in the Data Supplement
that accompanies the online version of this Technical Brief
Comparison with previous gene expression analysis of
CRC that used tissue from the tumor as the source of RNA
revealed that 4 of our differentially upregulated genes,
PSAM3, RANBP1, GCLC, and KIAA0101, had been previ-
ously identified in 2 different studies as upregulated in
CRC (12, 13).
We then performed quantitative RT-PCR (Q-RT-PCR)
on the same samples to analyze expression for a subset of
4 randomly selected genes, KIAA0101, UBE2D3, EPAS1,
and DDX46, from the list of 40 differentially upregulated
genes. Two of them were validated by PCR as signifi-
cantly upregulated in the CRC samples (KIAA0101 and
UBE2D3; both with Kruskal–Wallis (KW) P value ?0.05),
one showed clearly increased expression although with
lower statistical significance (EPAS1; KW P value ?
0.098), and the last one could not be validated (DDX46; P
value ? 0.96) (Table 1 and Supplementary Fig. 1B in the
online Data Supplement).
One of these 4 genes, KIAA0101, also known as p15PAF,
has been previously identified as a commonly overex-
pressed gene in a variety of solid tumors by 2 indepen-
dent groups who used large-scale metaanalysis of cancer
DNA microarray data (14, 15). EPAS1 encodes hypoxia-
inducible factor 2-alpha (HIF2?), an important angiogenic
factor whose high expression in CRC has been shown to
play an important role in tumor progression and to
possess prognostic value (16). UBE2D3, another of our
selected markers, encodes a ubiquitin-conjugating en-
zyme, also known as UBCH5C, involved in the regulated
degradation of important cellular factors such as the
tumor suppressor p53 and the NF?B regulator, I?B?
(17, 18). Finally, DDX46 encodes a member of the DEAD
box protein family that has putative helicase activity and
is involved in pre-mRNA splicing as part of the 17S U2
small nuclear ribonucleoprotein complex.
To test the consistency of the detection of these genes in
the blood of CRC patients, we analyzed their expression
by Q-RT-PCR on a set of 29 new CRC plasma samples and
36 healthy donor samples, different from the ones that
were part of the microarray study. With this external set
we verified the increased expression of 2 of the markers,
although only 1 of them (EPAS1, KW P value ?0.05) was
significantly higher in CRC patients than in healthy
Clinical Chemistry 53, No. 10, 2007
donors. The other marker (UBE2D3, KW P ? 0.09),
although its mean value was increased, had lower statis-
tical significance (Table 1 and Supplementary Fig. 2A in
the online Data Supplement). Finally, the overall analysis
of the above markers by Q-RT-PCR across all the samples
used in this study, i.e., the internal and the external sets
together, showed a clear and statistically significant (KW
P ?0.05) increase of KIAA0101, UBE2D3, and EPAS1 in
the plasma of CRC patients compared to healthy individ-
uals (Fig. 1A).
To test the power of the identified markers in enabling
differentiation of the tumor from the healthy condition,
we measured the levels of these markers in the plasma of
CRC patients (n ? 11), from whom it was possible to
obtain blood samples after surgical removal of the tumor.
Importantly, EPAS1 and UBE2D3 were significantly de-
creased in the postsurgery samples compared to the
presurgery samples from the same patients (Table 1 and
Supplementary Fig. 2B in the online Data Supplement).
To further explore the discriminating power of the
markers identified in this study, we applied a supervised
learning algorithm to the Q-RT-PCR dataset, excluding
the pre- and postsurgery data and using the resulting
dataset as the training set. Support vector machine (SVM)
analysis with leave-one-out cross-validation of this train-
ing set using the 3 validated genes UBE2D3, EPAS1, and
KIAA0101, showed that they enabled classification of up
to 71% of the training samples correctly (52 of 73 samples)
(see Fig. 1B). Use of only 2 markers, UBE2D3 and EPAS1,
did not improve scores obtained with the 3 genes to-
gether. Using the model generated by SVM with the 3
markers, we performed class prediction on a test set
composed of the Q-RT-PCR data derived from the pre-
and postsurgery group. In this way, we classified 77% of
the samples correctly (17 of 22 samples), i.e., presurgery
samples were classified as CRC and postsurgery samples
were classified as normal (Fig. 1B). The misidentified
samples were 3 presurgery samples that were wrongly
Fig. 1. Q-RT-PCR analysis of CRC markers in plasma.
(A), analysis of the mRNA concentrations of the selected markers in plasma of the complete series (internal plus external, see Table 1) of normal (N) and CRC patients
(CRC) samples used throughout the study (the number of samples in each case is shown below the class label). Genes for which a statistically significant difference
(P ?0.05) was shown by KW test are marked with (*). (B), class prediction using Q-RT-PCR expression data of our markers. A SVM algorithm with lineal kernel was
applied to construct a model for class prediction using Q-RT-PCR expression data for UBE2D3, EPAS1, and KIAA0101, or UBE2D3 and EPAS1 together, for all the
samples used in this study except for the pre- and postsurgery samples (u: training set). The models generated were used to classify the pre- and postsurgery samples
(f: test set). The graph shows the percentage of correctly classified samples in each case. Correctly classified and total number of samples are shown on top of each
classified as normal and 2 postsurgery samples assigned Download full-text
to the CRC group.
Simultaneous monitoring of the expression of numer-
ous genes by DNA microarrays provides a powerful tool
in medical research, but the widespread clinical applica-
tion of DNA microarrays is hindered by the need for
sample collection directly from the tumors. Analysis of
circulating RNA in the plasma circumvents this limita-
tion, making sample collection easy and reproducible,
and allowing for reiterative extractions during treatment
response. This proof-of-concept study demonstrates the
feasibility of such an approach. Our results provide an
example of the power of plasma RNA analysis to differ-
entiate tumor from the healthy condition in a clinical
setting. We observed that some of our markers, present at
high concentrations in CRC, returned to normal after
surgical removal of the tumor. Furthermore, class predic-
tion using SVM classified the presurgery samples as
members of the CRC group and the postsurgery samples
as part of the normal group.
On the basis of our results, large-scale gene expression
profiling of a large number of samples should yield
candidate markers of potential diagnostic and prognostic
Grant/funding support: This work was mainly funded by
Grant FUGEDAD from the Spanish Ministry of Education
and Science (MEC) (to M.S., F.B., and A.Z.). Additional
support was obtained for the laboratory of M.S. from the
CNIO, the MEC, and the European Union (INTACT and
PROTEOMAGE). The laboratory of F.B. was also funded by
MEC and Fundacion Mutua Madrilen ˜a.
Financial disclosures: None declared.
Acknowledgments: We thank Joaquin Dopazo and Ignacio
Medina from Centro de Investigacio ´n Prı ´ncipe Felipe, Valen-
cia, for their excellent support with class prediction.
1. Stroun M, Anker P, Lyautey J, Lederrey C, Maurice PA. Isolation and
characterization of DNA from the plasma of cancer patients. Eur J Cancer
Clin Oncol 1987;23:707–12.
2. Chen XQ, Stroun M, Magnenat JL, Nicod LP, Kurt AM, Lyautey J, et al.
Microsatellite alterations in plasma DNA of small cell lung cancer patients.
Nat Med 1996;2:1033–5.
3. Silva JM, Dominguez G, Garcia JM, Gonzalez R, Villanueva MJ, Navarro F, et
al. Presence of tumor DNA in plasma of breast cancer patients: clinicopath-
ological correlations. Cancer Res 1999;59:3251–6.
4. Silva JM, Dominguez G, Villanueva MJ, Gonzalez R, Garcia JM, Corbacho C,
et al. Aberrant DNA methylation of the p16INK4a gene in plasma DNA of
breast cancer patients. Br J Cancer 1999;80:1262–4.
5. Silva JM, Gonzalez R, Dominguez G, Garcia JM, Espana P, Bonilla F. TP53
gene mutations in plasma DNA of cancer patients. Genes Chromosomes
6. Chen XQ, Bonnefoi H, Pelte MF, Lyautey J, Lederrey C, Movarekhi S, et al.
Telomerase RNA as a detection marker in the serum of breast cancer
patients. Clin Cancer Res 2000;6:3823–6.
7. Hasselmann DO, Rappl G, Rossler M, Ugurel S, Tilgen W, Reinhold U.
Detection of tumor-associated circulating mRNA in serum, plasma and blood
cells from patients with disseminated malignant melanoma. Oncol Rep
8. Kopreski MS, Benko FA, Gocke CD. Circulating RNA as a tumor marker:
detection of 5T4 mRNA in breast and lung cancer patient serum. Ann N Y
Acad Sci 2001;945:172–8.
9. Kopreski MS, Benko FA, Kwak LW, Gocke CD. Detection of tumor messenger
RNA in the serum of patients with malignant melanoma. Clin Cancer Res
10. Silva JM, Dominguez G, Silva J, Garcia JM, Sanchez A, Rodriguez O, et al.
Detection of epithelial messenger RNA in the plasma of breast cancer
patients is associated with poor prognosis tumor characteristics. Clin
Cancer Res 2001;7:2821–5.
11. El-Hefnawy T, Raja S, Kelly L, Bigbee WL, Kirkwood JM, Luketich JD, et al.
Characterization of amplifiable, circulating RNA in plasma and its potential
as a tool for cancer diagnostics. Clin Chem 2004;50:564–73.
12. Alon U, Barkai N, Notterman DA, Gish K, Ybarra S, Mack D, et al. Broad
patterns of gene expression revealed by clustering analysis of tumor and
normal colon tissues probed by oligonucleotide arrays. Proc Natl Acad Sci
U S A 1999;96:6745–50.
13. Notterman DA, Alon U, Sierk AJ, Levine AJ. Transcriptional gene expression
profiles of colorectal adenoma, adenocarcinoma, and normal tissue exam-
ined by oligonucleotide arrays. Cancer Res 2001;61:3124–30.
14. Pilarsky C, Wenzig M, Specht T, Saeger HD, Grutzmann R. Identification and
validation of commonly overexpressed genes in solid tumors by comparison
of microarray data. Neoplasia 2004;6:744–50.
15. Rhodes DR, Yu J, Shanker K, Deshpande N, Varambally R, Ghosh D, et al.
Large-scale meta-analysis of cancer microarray data identifies common
transcriptional profiles of neoplastic transformation and progression. Proc
Natl Acad Sci U S A 2004;101:9309–14.
16. Yoshimura H, Dhar DK, Kohno H, Kubota H, Fujii T, Ueda S, et al. Prognostic
impact of hypoxia-inducible factors 1alpha and 2alpha in colorectal cancer
patients: correlation with tumor angiogenesis and cyclooxygenase-2 expres-
sion. Clin Cancer Res 2004;10:8554–60.
17. Saville MK, Sparks A, Xirodimas DP, Wardrop J, Stevenson LF, Bourdon JC,
et al. Regulation of p53 by the ubiquitin-conjugating enzymes UbcH5B/C in
vivo. J Biol Chem 2004;279:42169–81.
18. Gonen H, Bercovich B, Orian A, Carrano A, Takizawa C, Yamanaka K, et al.
Identification of the ubiquitin carrier proteins, E2s, involved in signal-induced
conjugation and subsequent degradation of IkappaBalpha. J Biol Chem
Table 1. Summary of expression values and statistical
analysis for the selected CRC markers across the internal,
external, and pre- and postsurgery groups (n ? number of
samples in each case).
4.1 Fold change microarray
Fold change Q-PCR
Normal (n) CRC (n)
Fold change Q-PCR
Normal (n) CRC (n)
Fold change Q-PCR
1.5 2.8 1.61.1
(35) (28) (35) (28)
0.5 0.10.6 1.2
(11) (11) (11) (11) (11) (11) (11) (11)
Clinical Chemistry 53, No. 10, 2007