Download Ofloxacin as a Reference Marker in Hair of Various Colors

Journal of Analytical Toxicology, Vol. 27, April 2003
Technical Note I
Ofloxacin as a Reference Marker in Hair of
Various Colors
Diana G. Wilkins 1, Atsuhiro Mizuno 2, Chad R. Borges 1, Matthew H. Slawson 1, and Douglas E. Rollins 1
1Centerfor Human Toxicology, Departmentof Pharmacologyand Toxicology, Room 490 Biomedical PolymersResearch
Building, Universityof Utah, 20 South 2030 East,Salt Lake City, Utah 84112 and 2Phase 1 and Clinical Pharmacology
Department, GlaxoSmithkline K.K., 6-15, Sendagaya4-chome, Shibuyaku, Tokyo 151-8566,Japan
[Abstract
It has been proposed that administration of a reliable marker
substance to human subjects may enhance the ability to identify
drug use and treatment compliance in drug treatment programs.
The goal of this study was to determine if an oral dose of the
antibiotic ofloxacin (OFLX) could be used as a "marker" substance
to establish reference points with respect to time in hair of various
colors. Male and female subjects (n = 32) between 18 and 40 years
of age received 800 mg of OFLX as a divided oral dose on a single
day. Subjects were restricted from cutting their hair or performing
chemical treatments. Hair was collected (by cutting) before, and at
weeks 4, 5, 6, and 7 after drug administration. Subjects were
classified as having black (n = 5), brown (n = 13), blonde (n = 8), or
red (n = 6) hair. Hair was segmented into 3.0-cm segments prior to
digestion, extraction, and analysis by high-pressure liquid
chromatography (HPLC). At 7 weeks, the mean OLFX
concentrations (• 1 SD) in the first 3.0 cm of hair closest to the
scalp were as follows: 30.6 • 8.5 ng/mg (black), 6.0 • 1.8 ng/mg
(brown), 3.5 • 1.6 ng/mg (blonde), and 1.4 • 0.3 ng/mg (red).
A similar pattern was found in hair collected at weeks 4-6.
Quantitative eumelanin (EUM) hair concentrations for each subject
were also determined for each subject via HPLC. A strong
relationship between OFLX concentration at 7 weeks and EUM was
noted (r2 adjusted = 0.728; p < 0.001). In six subjects, we also
determined the intrasubject variability of OFLX incorporation into
individual hair strands. Four strands from each subject were
segmented into 2-mm segments and analyzed. OFLX appeared in
segments #1-#10 at week 5 (the first centimeter of hair). OFLX
appeared in segments #2-#20 at week 7 (the first and second
centimeter of hair). The maximum OFLX concentration (the "band"
of drug) and location was then determined for each strand. The
maximum OFLX concentration was measured in segments #2-#5 at
week 5 for all subjects (within the first centimeter of hair length).
The maximum OFLX concentration was measured in segments
#3-#8 at week 7 (within the first and second centimeter of hair).
This was consistent with a growth rate of less than 1.0 cm/month,
although considerable intersuhject variability was found. No
significant axial diffusion of OFLX along the hair shaft beyond the
first 3.0 cm of hair was noted. Despite a strong effect of hair color,
these data suggest that OFLX may be a suitable marker substance
for hair, allowing a subject to serve as their own "control". Future
studies will explore whether drug use, treatment compliance, or
recidivism in clinical drug-abuse studies can be determined with
the aid of OFLX.
Introduction
The analysis of hair may be a useful adjunct specimen to
plasma and urine for monitoring compliance and recidivism in
drug treatment programs. It has been suggested that hair may
serve as an historical record, or diary, of drug exposure (1,2).
Improved methods for determining patient compliance over
time are necessary because drug concentrations in plasma,
urine, and saliva often reflect only the dosage taken within the
last several days prior to sampling. Our ability to accurately interpret drug concentrations in hair, however, is uncertain, and
there is conflicting evidence regarding the utility of hair analysis for drug monitoring (3-31). There is evidence from small,
controlled studies in humans and animals that the variability in
measured hair concentrations of drug, given an equal dose, is affected by hair pigmentation (32--47). However, at least two reports have suggested that hair color does not play a role in the
outcome of hair drug testing (48,49). These data are not necessarily contradictory, as the outcome of the drug test may be
dependent upon the cutoff concentrations used to identify a positive versus a negative specimen. Until these issues are clarified,
the ability to interpret trends in quantitative hair data over
time is limited. It has been suggested that the use of "marker"
compounds may enhance the assessment of treatment compliance and recidivism by establishing known reference points
(with respect to time) in hair, despite interindividual variability
(45,50,51). These marker compounds may permit the assessment of illicit or therapeutic drug exposure in subjects participating in clinical treatment programs.
Ofloxacin (OFLX) is a fluorinated carboxyquinolone antibiotic
with approximately 98% bioavailability after oral administration. Previous research demonstrated that OFLX could be detected in hair after therapeutic use and multiple-dose administration (50-55). Uemtasu et al. (50) first suggested the possible
Reproduction(photocopying)of editorialcontentof thisjournalis prohibitedwithoutpublisher'spermission.
149
Journal of Analytical Toxicology, Vol. 27, April 2003
use of ofloxacin as an index of exposure. In that study, the
OFLX content was measured in hair collected from 14 subjects
known to have taken OLFXfor treatment of bacterial infection.
Although exact doses and duration of therapy varied among
these subjects, it did demonstrate that OFLX moved outwards
along the hair shaft after administration. The authors suggested that the compound could be useful in the future for
testing patient compliance. Further studies in humans and albino and pigmented rats by this same group suggested that
OFLX is excreted in a dose-dependent manner and that the
mechanism of excretion is closely linked with the presence of
melanin (56). However, only four human subjects with
black/grizzled hair were included in the study, and it was unclear to what extent OFLX would be incorporated in hair of
other colors.
As suggested by these early studies, administration of OFLX
(as a marker) at specific, but limited, time points during drug
treatment may establish a frame of reference with which to
determine concomitant use of other drugs. The purpose of this
study was to determine if the OFLXcould be used as a "marker'
substance to establish reference points with respect to time in
hair of various colors. First, we hypothesized that OFLXcan be
readily detected in human hair after single day of administration, regardless of hair color. Second, we hypothesized that
OFLXwill move along the hair shaft in a pattern consistent with
natural hair growth (not simple diffusion along the shaft).
Materials and Methods
Chemicals and reagents
OFLX, 40% tetrabutylammonium hydroxide solution, and
orthophosphoric acid were obtained from Sigma (St. Louis,
MO). (R)-9-Fluoro-2,3-dihydro-3-methyl- 10-(4-ethyl-l-piperazinyl)-7-oxo-7-hydroxy-pyridol[1,2,3,-de][1,4]benzoxazine-6carboxylic acid (DS-4632) for use as an internal standard was
graciously donated by Daiich Pharmaceutical Co. (Tokyo,
Japan). Chloroform and hydrochloric acid were obtained from
Burdick and Jackson (Muskegon, MI) and Mallinckrodt Chemical (St. Louis, MO), respectively.
Standards and solutions
Stock solutions of the OFLX (1 mg/mL) and DS-4632 internal standard (0.5 mg/mL) were prepared in methanol and
stored at -4~ Dailyworking solutions of OFLXwere prepared
in methanol at 10.0 ng/mL, 100.0 ng/mL, and 1.0 IJg/mL.A serial dilution of the DS-4632 stock solution was prepared, in
methanol, to a final concentration of 500.0 ng/mL. Working solutions of OFLX were used to fortify human hair at final concentrations of 0.5, 1.0, 3.0, 5.0, 10.0, 20.0, 30.0, and 50.0 ng/mg
(standard curves). Positive quality control specimens (5 and
30 ng/mg, final concentrations) for verification of accuracy and
precision were analyzed in duplicate with each analytical batch.
Separate stock solutions of OFLXwere prepared from reference
materials for standards and quality control samples.
Study protocol
Written informed consent approved by the University of Utah
150
was obtained from all study participants. All enrolled subjects
claimed no prior use of OFLX within at least the previous six
months. Subjects were housed in the Clinical Research Center
at the University of Utah Health Sciences Center at the time of
OFLX administration. Males and females (n = 32) between 18
and 40 years of age received 800 mg of OFLXas an equally divided oral dose at 0800 and 2000 h. Plasma was collected 12 h
after the last dose. Hair was collected (by cutting) at the vertex
region of the scalp before and at weeks 4, 5, 6, and 7 after drug
administration. Subjects were visuallyclassifiedas having black
(n = 5), brown (n = 13), blonde (n = 8), or red (n = 6) hair. Subjects were restricted from cutting their hair or performing
chemical treatments for the duration of the study. Hair was
stored at -20~ until analysis.
Analytical procedures
Ofloxacin in hair of different colors. All hair strands from
each subject were individually aligned root-to-tip. Then, approximately 10 hairs from each subject were segmented into
3.0-cm segments. Thus, each 3.0-cm segment (sample) consisted of approximately 10 hairs.
Ofloxacin concentrations in hair were determined by a modification of the method of Mizuno et al. (57). Internal standard
(25 ng/mg DS-4632) was added to a l-rag samplefrom each segment. A Cahn TA4100 electrobalance was used to weigh specimens (+ 1.0% tolerance). Standards and quality control specimens containing hair fortified with known amounts of OFLX
were included with all assays as described. Samples were completely dissolved in 2 mL of 1N NaOH at 70~ for 20 rain. After
cooling to room temperature, the pH was adjusted to 9.0 with
several drops of 6N HC1and 1 mL of phosphate buffer (pH 9.0).
Digest solutions were extracted with Bond-Elut CertifyTM solidphase extraction columns. Columns were prewashed with distilled water and methanol, and the digest solution was applied
to the column, followedby column rinses with 3 mL of distilled
water, 2 mL of 0.1M acetate buffer (pH 4.0), and 5 mL of
methanol. OFLXwas then eluted with 5 mL of methylene chloride/isopropanol (80:20) containing 2% ammonium hydroxide.
Eluates were evaporated to dryness at 40~ in a water bath.
Dried extracts were reconstituted in 7501JL of mobile phase; 50
laL was injected for high-pressure liquid chromatography
(HPLC) analysis. The limit of quantitation for OLFX (hair) in
this procedure was 0.5 ng/mg of hair.
Ofloxacin distribution in individual hair strands and plasma.
Two different methods were used for the determination of
ofloxacin in "intact" versus "individual" hairs to increase our
confidence and ability to compare data to previous research
(both methods had been previously published and validated).
For individual single strands, hairs were individually aligned
root-to-tip and each strand segmented into 2-mm segments.
OFLXconcentrations in hair were determined by a modification
of the method of Mizuno et al. (57). Internal standard (25 ng
DS-4632) was then added to each 2-mm segment. Standards
and quality control specimens containing hair fortified with
known amounts of OFLXwere included with all assays. Samples
were completely dissolved in 0.5 mL of 1N NaOH at 70~ for 10
rain. After cooling to room temperature, the pH was adjusted to
7.0 with 0.5mL of 1N HCl and 1 mL of phosphate buffer (pH
Journal of Analytical Toxicology, Vol. 27, April 2003
7.0). Digestsolutions were extracted with chloroform (5 mL) for
20 rnin, centrifuged at 1670 x g for 10 rain, and the organic
phase collected. Extracts were evaporated to dryness at 40~ in
a water bath. Dried extracts were reconstituted in 150 I~Lof mobile phase; 50 IJLwas injected for HPLC analysis. Plasma specimens were extracted and analyzed similarly,with the following
exceptions. Internal standard (500 ng) was added to 100 iJL of
plasma, 1 mL of 0.5M phosphate buffer (pH 7.0) was then added
and OFLXextracted into chloroform. The limits of quantitation
for OFLXin this procedure were 0.2 ng/2 mm hair segment or
0.2 ng/mL plasma.
HPLr analysis ofOFLX. HPLCanalyses were performed on a
Waters (Milford, MA) 600E multisolvent delivery system
equipped with a Waters 600 controller, Waters 717plus autosampler, and model 474 scanning fluorescence detector. Reconstituted extracts were injected onto a SymmetryTM 5 tim C18
250 x 4.6-ram column. The mobile phase consisted of acetonitrile/0.025M orthophosphoric acid adjusted to pH 3.0 with 40%
tetrabutylammonium hydroxide solution (6.5:93.5, v/v). The
flow rate and excitation and emission wavelengths were 1.0
mL/min, 290 nm, and 490 nm, respectively. Peak-height ratios
of OFLX to DS-4632 internal standard were compared to a
standard curve made from a series of standards extracted concurrently with the specimens.
Eumelanin in hair. The degradation products pyrrole-2,3-dicarboxylic acid (PDCA) and pyrrole-2,3,5-tricarboxylic acid
(PTCA) of eumelanin were measured by a modification of the
method of Ito and Wakamatsu (58). Five milligrams of hair
was cut into 1-2-ram pieces and degraded in 1 mL of 0.5M
NaOH containing 80 IJL of 3% H202 by heating in a boiling
water bath for 20 rain. Pthalic acid (40 nmol) was added as internal standard prior to degradation. Complete details of the
modified method used for this study have been previously published (59).
LC conditions for eumelanin analysis. Quantitation of eumelanin degradation products (PDCAand PTCA)and internal
standard was carried out with a Waters 600E HPLC system
with UVdetection at 280 nm. Samples (100 I~L)were injected
onto a Phenomenex (Torrance, CA) Luna 5 IJm C18 250 • 4.6mrn column at a temperature of 55~ The mobile phase consisted of 0.01M potassium phosphate buffer (pH 2.1) and
methanol at a flow rate of 0.8 mL/min under the following gradient: 98%/2% aqueous/organic ramped to 40%/60%
aqueous/organic over 14 rain, held for 6 rain, and returned to
Table I. Ofloxacin and Eumelanin in Hair
Nalural
Hair Color
Gender
Mean Ofloxacin
Hair Conc.(TW)
(Self.Reporl)
(M/F)
(ng/mg)
Red
Blonde
Brown
Black
2 Males/4 Females
6 Males/2 Females
7 Males/6 Females
2 Males/3 Females
1.49 (+ 0.38)*
3.51 (• 1.68)*
6.03 (• 1.81)*
30.64(• 8.55)*
MeanPlasma
Conc. at 12 h
(rig/mr)
1394.0 (•
1304.9 (•
1270.3 (•
1291.0 (•
* Significantly different from all other hair colors at p < 0.05.
Significantly different from brown and black hair only at p < 0.05.
649.8)
296.5)
298.4)
238.8)
98%/2% over 5 min. Peak-height ratios of PDCAand PTCAto
internal standard were compared to a standard curve made
from pure PDCAand PTCAstandards subjected to alkaline hydrogen peroxide degradation.
Statistical analysis
Simple linear regression analysis, ANOVA,and tests of significance were performed with SPS$ (Statistical Package for the
Social Sciences, version 6.1, Chicago, IL) and DataDesk (version
4.0, New York, NY).
Results and Discussion
A suitable marker substance for use in clinical studies of
drug incorporation into hair would be one that is detectable
after a single dose, is present in high concentrations, and has
minimal or no adverse pharmacologic effects on the individual.
It should also move along the hair shaft in a predictable pattern,
remaining as a tight "band" to permit establishment of a
"window of time" within the hair. Ideally, the marker should
also be one that has no potential for passive exposure from the
environment and limited abuse liability. OFLXwas selected for
this study because it appeared to meet most of these criteria and
it had been previously reported that administration of multiple
oral doses OFLXto humans was associated with large measurable concentrations in hair (50-55). These studies also strongly
suggested that OFLXwould be a useful time marker for both
clinical and forensic purposes.
In our current study, we proposed to investigate whether
OFLXcould be detected in human hair after 1-day oral dose administration and whether its detection within the hair shaft followed a pattern consistent with normal hair growth. In particular, we were interested in whether the OFLX marker moved
along the hair shaft as a '%and" of drug, allowing us to estimate
the time of original exposure. Baseline hair specimens were
collected just prior to oral administration of OFLX. OFLXwas
not detected in these pre-dose hair specimens, indicating that
the subjects had either not taken the drug, or had not previously
ingested a sufficient amount to produce a detectable concentration. At 7 weeks, the hair concentrations of OFLX for all
subjects ranged from a low of 0.84 to a high of 44.70 ng/mg in
the first 3 cm of hair closest to the scalp. Assuming an average
hair growth rate of approximately 1.0
era/month, we expected that OFLX would not
be detected beyond the first 3-cm of hair during
this study. Consistent with this expectation,
MeanEumelanin
Concentration OFLX was not detected in any segments beyond 3 cm during the study.
(pg/mg)
Although OFLXwas readily measured in hair,
a large interindividual variability in measured
1.68 (• 0.69)
OFLXconcentration was observed. Previous ev2.25 (+ 0.33)*
3.51 (• 0.95)*
idence has suggested that that hair pigmenta10.92 (• 3.6)*
tion may be an important factor in the incorporation of some drugs into human hair
(32-47). These studies support the hypothesis
that drugs that retain a positive "charge" at
151
Journal of Analytical Toxicology, Vol. 27, April 2003
physiologic pH may be preferentially incorporated into melanincontaining cells of the hair shaft. Of relevance to our current
study, Uematsu et al. (56) demonstrated significantly greater
OFLX concentrations in black versus white hairs from four
subjects described as having "grizzled" hair. To determine
whether differences in hair pigmentation contributed to the
variability in OFLX concentrations in our study, hair from each
subject was classified into one of four hair-color categories
(black, brown, blonde, and red). The initial classification was
made on the basis of self-report and confirmed by visual observation by laboratory staff. A summary of hair concentration
data and hair color is presented in Table I. At 7 weeks, the mean
OLFX concentrations (• 1 SD) in the first 3 cm of hair closest
to the scalp were as follows: 30.6 • 8.5 ng/mg (black), 6.0 • 1.8
ng/mg (brown), 3.5 • 1.6 ng/mg (blonde), and 1.4 • 0.3 (red)
ng/mg. Differences were statistically significant (ANOVA,p <
y2 (adjusted) = 0.728
p < 0.001
{
A
a SubJect
,
an uonnoz
.:
NILII~
O
.
. i
-' . . . .
5
10
15
Eumelanln concentration
(IJglmg)
20
Figure 1. Ofloxacin versus eumelanin concentrations. The ofloxacin
concentration for hair from each subject (collected 7 weeks after OFLX
administration) versus their respective eumelanin concentration.
9 .... ..,, ~.~,,,~.,~ ,~.:,~,,~ ~.....
[
..
.
-
,
2
4
6
8
10
12
14
16
18
20
LocaUon from hair root
(ran)
Figure 2. Ofloxacin distribution in 2-mm hair segments. Four individual
hair strands (collected at week 7) from six subjects were each segmented
into 2-mm lengths and analyzed for OFLX. Error bars along the x-axis reflect the standard deviation of the range of location (ram) from the hair
root in which maximum OFLX concentrations were detected. Error bars
along the y-axis reflect the standard deviation of the range of maximum
OFLX concentration measured in the four hair strands.
152
Table II. Variation in Growth Rate of Individual Hair
Strands
Subject ID
.
,,p
0
0.05, DataDesk) between all hair-color groups.
We then evaluated whether differences in plasma concentrations accounted for the interindividual variability in OFLX hair
concentrations. Plasma OFLX concentrations (12 h) were not
significantly different between subjects with different hair colors
(see Table I). The mean OFLX plasma concentrations (• 1 SD)
at 12 h for subjects with black, brown, blonde, and red hair were
1394.0 • 649.8 ng/mL, 1304.9 • 296.5 ng/mL, 1270.3 • 298.4
ng/mL, and 1291.0 • 238.8 ng/mL, respectively. Based on the
12-h data, it is unlikely that differences in peak plasma concentrations of OFLX explain the variability in hair concentrations.
Because the classification of hair color into categories is
largely a subjective assessment, we implemented a more objective measure of hair color with the use of quantitative eumelanin concentrations. Our laboratory has previously developed an improved quantitative procedure for the determination
of eumelanin concentrations in human hair (59). The method
was used for analysis of eumelanin in hair specimens in this
current OFLX study (see Table I). The mean measured eumelanin concentrations (• 1 SD) for subjects with black, brown,
blonde, and red hair were 10.92 + 3.6 I~g/mg,3.51 • 0.95 IJg/mg,
2.25 • 0.33 IJg/rng, and 1.68 • 0.69 IJg/mg, respectively. Differences were statistically significant between black hair and all
other colors (ANOVA,p < 0.05, DataDesk), as well as brown hair
versus red and blonde hair. There was no statistical difference in
eumelanin concentration between blonde and red hair. These
data are consistent with trends in eumelanin content previously reported for black, brown, and blonde hair based on the
PTCA content of human hair, as determined by Ito and Fujita
(60). The data trends are also consistent for all four hair colors
with the spectrophotometric eumelanin determinations of
human hair performed by Ito et al. (61).
A positive relationship between OFLX concentration at 7
Location of Maximum
OFLX Concentration
(Range of Segments)*
#1
(Female;Asian; Black Hair)
#2
(Mate; African-American; Black Hair)
#3
(Male; Caucasian;Brown Hair)
#4
(Male; Caucasian;Brown Hair)
#5
(Male; Caucasian;Blonde Hair)
#6
(Female; Caucasian;Blonde Hair)
Estimated
Growth Rate
(cm/month) f
#5-#8
0.61-0.98
#4-#5
0.49-0.61
#5-#7
0.61-0.85
#5-#7
0.61-0.85
#4-#8
0.49-0.98
#3-#5
0.37-0.61
* The range of segments in which the maximum OFLX concentration was
measured for four hair strands per subject.
t The estimated hair growth rate, based on the location of maximum OFLX
concentration and the time since OFLX administration (49 days; see text for
details).
Journal of Analytical Toxicology, Vol. 27, April 2003
weeks and EUMwas noted (Figure 1). The independent variable
(EUM concentration) accounted for approximately 72.8% of
the variance in the dependent variable (OFLX, r z adjusted =
0.728,p < 0.001). The adjusted r z was selected to prevent an "inflation" of r from overfitting the data. Therefore, the concentration of eumeianin in hair does appear to be associated with
the amount of OFLX incorporated into hair. Subjects with the
darkest hair (confirmed by eumelanin concentrations) incorporated more drug than individuals with lighter-colored hair.
The effectof gender and age were also evaluated;however,no association was observed for these two variables with OFLXconcentration in hair.
Our data also indicate, however, that eumelanin is not the
only factorinvolvedOFLX incorporation because all of the variability in OFLX concentration could not be explained by the
subject's eumelanin concentration. Some of the variability may
be explained by limitations of the analytical procedure for eumelanin itself. The results of the analytical procedure are based
on the assumption that tyrosinase-produced synthetic melanins
are accurate and reproducible models of in vivo biological
melanins. As the exact three-dimensional structure of eumelanin is unknown (62), the effect of potential differences between synthetic and biological melanins in this study cannot be
determined.
Also, it is highly probable that there are other regions of
drug-binding to hair components that have yet to be elucidated. It is known that basic drugs can be incorporated into
non-pigmentedhair, albeit at lower concentrations than in pigmented hair (33--44). Despite basic structural similarities
among all hair types regardless of hair color, there do appear to
be some differencesin the chemical and physical characteristics
of ethnic hair types together with considerable intra-ethnic
variation (63,64). Differences include such factors as the diameter of the hair shaft, degree of medullation, curvature of the
hair shaft, crimp, cross-sectional shape, protein content, and
follicular form, to name a few. It is possible that differences in
the ultra-structure, morphology, and protein content and structure between hair types may play a lesser role in the binding of
OFLX to hair.
In six subjects, we determined the intrasubject variability of
OFLX incorporation into individual hair strands (Figure 2).
Four entire hair strands from each subject were segmented
into 2-mm segments, and each segment was analyzed for OFLX.
The maximum OFLXconcentration (the '~and" of drug) and location were then determined for each strand. The mean maximum OFLXconcentration in the 2-ram hair segments (• 1 SD)
ranged from 0.55 (• 0.25) to 6.74 (• 1.55) ng/mg. The maximum OFLXconcentration was measured in segments #2--#5 at
week 5 for the six subjects (within the first centimeter of hair
length). The maximum OFLXconcentration was measured in
segments #3--#8 at week 7 (within the first and second centimeter of hair). Assuming a hair growth rate of 1.0 cm/month,
and noting that the last hair specimen was collected 7 weeks
after OFLXadministration, we hypothesized that OFLXwould
be detected only in the first and second centimeters (the 20-mm
closest to the scalp) of hair of most subjects. As expected, our
data demonstrate that no significant axial diffusion of OFLX
along the hair shaft occurred during the period of time en-
compassed by this study. This suggests that normal hygienic
practices (e.g., regular hair washing) do not result in movement
of the OFLXalong the hair shaft.
We also explored whether the time of OFLX administration
could be reasonably extrapolated from the location of the maximum measured OFLX concentration at week 7 (Table II). The
range of location for the mean maximum OFLXconcentration
for each of the six subjects was as follows: Subject #1 (sagments #5--#8/~10-16 mm); Subject #2 (segments #4-#8/~8--10
ram), Subject #3 (segments #5-#7/~10-14 turn), Subject #4
(segments #5-#7/~10-14 ram), Subject #5 (segments
#4--#8/~8-16 rnm), and Subject #6 (segments #3--#5/~6-10
ram). If an estimated hair growth rate of 1.0 cm/month (or
~0.033 crrgdayfor an average of 30 days) was assumed, then the
maximum OFLX concentration would be expected to be located at about 16 mm at 7 weeks (49 days after receiving the
dose). However,as shown in Table II, the estimated hair growth
rate for individual hair strands in these six subjects varied considerably and was generally less than 1.0 cm/month. For the six
subjects in this study, a 3-cm hair segment represents a much
longer 'Window of time" (> 3 months) than would be predicted
using an assumed 1.0 cm/month growth rate. The data from our
study suggest that it may not be possible to determine the use
of another drug to within a time frame of one month; however,
our study was not designed to specificallyaddress this question.
It should be also be noted that these findings are based on a limited number of subjects, as well as a limited number of hair
strands per subject. The wide range in growth rate estimates for
any single subject may be partly explained by the stage of hair
growth of the individual hair follicles during the period of drug
administration. For example, anagen-phase (actively growing)
hair may incorporate drug more readily than hair in catagen- or
telogen-phases of the hair growth cycle (3).
The variability in growth rate in hair specimens from this
study is inconsistent with that obtained in earlier studies of
multiple-dose OFLXadministration (51-53). These studies indicated an estimated mean hair growth rate (• 1 SD) of 1.12 •
0.11 cm/month. However, there are numerous differences between the study designs that may contribute to the differences
observed between the studies. These differencesinclude method
of hair collection (plucking versus cutting), inclusion of hair
bulb in hair length estimates, gender of individualsstudied, ethnicity, hair color, OFLXdose, and frequency of OFLXadministration. These discrepancies suggest that a more detailed examination of the variability in hair growth rates and the use of
OFLXas a reference marker must be explored. Such an assessment is necessary to enable the toxicologist to provide more accurate estimates of potential drug exposure or use. The authors are currently conducting a study with controlled
administration of OFLXand codeine, at various time intervals,
to determine whether OFLXcan establish reference points for
the time of codeine use within the hair shaft.
OFLXwas readily detected in the hair of all subjects, regardless of hair color, and did not diffuse along the hair shaft independent of natural hair growth. This demonstrates that OFt,X,
and perhaps other fluoroquinolone antimicrobial compounds,
can be useful "time markers" in hair under certain conditions.
As discussed previously, these compounds are attractive for use
153
Journal of Analytical Toxicology,Vol. 27, April 2003
as biomarkers because they are present in high concentrations
in hair, have minimal adverse pharmacologic effects, and have
limited abuse liability and potential for passive exposure from
the environment. However, OFLX is one of the most widely
used fluorquinolones for the treatment of infections (65). Phototoxicity and hypersensitivity reactions have been reported, although they are less likely to occur with single doses. Another
important limitation is that OFLX-resistant microorganisms
may emerge during fluoroquinolone therapy. Overuse and inappropriate use of fluoroquinolones can erode their clinical
utility for the treatment of future infections. Therefore, the
benefit of a single dose of OFLX, given to otherwise healthy individuals, at intervals of several weeks must be carefully weighed
against the potential risk(s) associated with antimicrobial use.
Ideally, the search for suitable alternatives for use in clinical hair
studies will continue.
Conclusions
OFLX was detected in human hair after a single therapeutic
dose to normal volunteers. Also, OFLX was only detected in the
first 3.0 cm closest to the scalp through week 7 of the study, indicating a lack of axial diffusion of OFLX along the hair shaft.
Hair color, as determined by the total eumelanin content (in
pg/mg), was associated with the incorporation of OFLX into
human hair. Despite this significant effect of hair color, these
data strongly suggest that OFLX is a suitable marker substance
for hair because it is readily detected in hair of all colors. The
use of compounds such as OFLX may permit the evaluation of
treatment compliance and recidivism in drug treatment programs, allowing a subject to serve as their own "control". Further studies are needed to determine whether OFLX can be
used as a marker substance so that determination of other drug
use (such as illicit drugs) can be determined.
Acknowledgments
This work was supported by National Institutes of Health
Grant DA09096.
References
1. P. Kintz. Clinical applications of hair analysis. In Drug Testing In
Hair, P. Kinlz, Ed. CRC Press, Boca Raton, FL, 1996, pp 267-277.
2. M. Harkey and G. Henderson. Hair analysis for drugs of abuse. In
Advances in Analytical Toxicology, Vol. 2, R. Baselt, Ed. Year
Book Medical Publishers, Chicago, IL, 1989, pp 298-329.
3. D. Rollins, D. Wilkins, S. Gygi, M. Slawson, and P. Nagasawa.
Testing for drugs of abuse in hair--experimental observations and
indications for future research. Forensic ScL Rev. 9:23-35 (1997).
4. D.E. Rollins, D.G. Wilkins, and G.G. Krueger. Codeine disposition
in human hair after single and multiple doses. Eur. J. Clin. Pharmacol. 50-"391-397 (1996).
5. Y. Nakahara, M. Shimamine, and K. Takahashi. Hair analysis for
154
drugs of abuse. III. Movement and stability of methoxyphenamine
along hair shaft with hair growth. J. Anal. Toxicol. 16:253-257
(1992).
6. A. Mizuno, T. Uematsu, A. Oshima, M. Nakamura, and M.
Nakashima. Analysis of nicotine content of hair for assessing individual cigarette-smoking behavior. Ther. Drug Monit. 15:99-104
(1993).
7. R. Sato, T. Uematsu, R. Sato, S. Yamaguchi, and M. Nakashima.
Human scalp hair as evidence of individual dosage history of
haloperidol: prospective study. Ther. Drug Monit. 11:686-691
(1989).
8. E. Cone. Testing human hair for drugs of abuse. I. Individual dose
and time profiles of morphine and codeine in plasma, saliva, urine
and beard compared to drug-induced effects on pupils and behavior. ]. Anal. Toxicol. 14:1-7 (1990).
9. I. Ishiyama, T. Nagai, and S. Toshida. Detection of basic drugs
(methamphetamine, antidepressants, and nicotine) from human
hair. ]. Forensic 5ci. 28:380-385 (1983).
10. M. Moeller, R Fey, and R. Wennig. Simultaneous determination of
drugs of abuse (opiates, cocaine, and amphetamine) in human hair
by GC/MS and its application to a methadone maintenance program. Forensic $ci. Int. 63:185-206 (1993).
11. H.A. Schroeder and A.P. Nason. Trace metals in human hair.
J. Invest. DermatoL 53:71-78 (1989).
12. K.M. Hambidge. Hair analysis: worthless for vitamins, limited for
minerals. Am. J. Clin. Nutr. 36:943-949 (1982).
13. J.M. McKenzie. Alteration of zinc and copper concentration of hair.
Am. ]. Clin. Nutr. 31:470-476 (1978).
14. S. Yamaguichi, H. Matsumoto, S. Kaku, M. Tateishi, and M. Shiramizu. Factors affecting the amount of mercury in human scalp
hair. Am. J. Pub. Health 65:484-488 (1975).
15. L. Kopito, R.K. Byers, and H. Schwachman. Lead in hair of children
with chronic lead poisoning. N. EngLJ. Med. 276:949-953 (1967).
16. K.M. Hambidge. Increase in hair copper concentration with increasing distance from the scalp. Am. J. Clin. Nutr. 26:1212-1215
(1973).
17. A. Baumgartner, R Jones, W. Baumgartner, and C Black. Radioimmunoassay of hair for determining opiate-abuse histories.
J. Nucl. Med. 20:748-752 (1979).
18. K. Puschel, P. Thomasch, and W. Arnold. Opiate levels in hair.
Forensic 5ci. Int. 21: 181-186 (1983).
19. M. Marigo, F. Tagliaro, C. Poiesi, S. Lafisca, and C. Neri. Determination of morphine in the hair of heroin addicts by high performance liquid chromatography with fluorometric detection.
J. Anal. Toxicol. 10." 158-161 (1986).
20. A. Baumgartner, R Jones, and W. Baumgartner. Detection of phencyclidine in hair. J. Forensic 5ci. 26:576-581 (1981).
21. O. Suzuki. Detection of methamphetamine and amphetamine in
a single human hair by GC/CI-MS. J. Forensic 5ci. 29:611-617
(1984).
22. R. Martz. The identification of cocaine in hair by GC/MS and
MS/MS. Crime Lab. Dig. 15:67-73 (1988).
23. G.L.Henderson, M.R. Harkey, C. Zhou, R.T.Jones, and R Jacob, III.
Incorporation of isotopically labeled cocaine and metabolites into
human hair. 1. Dose-response relationships. J. AnaL ToxicoL 20:
1-12 (1996).
24. T. Uematsu. Therapeutic drug monitoring in hair samples~principles and practice. Clin. Pharmacokinet. 25:83-87 (1993).
25. A. Mizuno, T. Uematsu, S. Gotoh, E. Katoh, and M. Nakashima.
The measurement of caffeine concentration in scalp hair as an indicator of liver function. J. Pharm. Pharmacol. 48:660-664 (1996).
26. A. Tracqui, R Krissig, R Kintz, A. Pouliques, and P. Mangin. Determination of amitriptyline in the hair of psychiatric patients.
Hum. Exp. ToxicoL 11:363-367 (1992).
27. J. Williams, RN. Patsatos, and J.F.Wilson. Hair analysis as a potential index of therapeutic compliance in the treatment of epilepsy.
Forensic 5ci. Int. 84:113-122 (1997).
28. A. Tracqui, R Kintz, and R Mangin. Hair analysis: a worthless tool
for therapeutic compliance monitoring. Forensic 5ci. Int. 70:
Journal of Analytical Toxicology,Vol. 27, April 2003
183-189 (1995).
29. F. Pragst, M. Rothe, and S. Thor. Structural and concentration effects on the deposition of tricyclic antidepressants in human hair.
Forensic $cL Int. 84:225-236 (1997).
30. O. Suzuki, H. Hattori, and M. Asano. Nails and hair as useful material for detection of methamphetamine or amphetamine abuse.
Forensic 5cL Int. 24:9-16 (1984).
31. U. Runne, F.R.Ochsendoff, K. Schmidt, and H.W. Raudonat. Sequential concentration of chloroquine in human hair correlates
with ingested dose and duration of therapy. Acta Derm. VenereoL
72:355-357 (1992).
32. H. Matsuno, T. Uematsu, and M. Nakashima. The measurement of
haloperidol and reduced haloperidol in hair as an index of dosage
history. Br. J. Clin. PharmacoL 29:187-194 (1990).
33. C.R.Borges, D.G. Wilkins, and D.E. Rollins. Amphetamine and Nacetylamphetamine incorporation into hair: an investigation of
the potential role of drug basicity in hair color bias. J. Anal ToxicoL 25:221-227 (2001).
34. M.H. Slawson, D.G. Wilkins, and D.E. Rollins. The incorporation
of drugs into hair: relationship of hair color and melanin concentration to phencyclidine incorporation. J. Anal. ToxicoL 22:
414-417 (1998).
35. D.G. Wilkins, A. Valdez, P.R. Nagasawa, S.P. Gygi, and D.E.
Rollins. Incorporation of drugs for the treatment of substance abuse
into pigmented and nonpigmented hair. J. Pharm. 5ci. 87(4):
435-440 (1998).
36. L. Potsch, G. Skopp, and M. Moeller. Influence of pigmentation on
the codeine content of hair fibers in guinea pigs. ]. Forensic 5cL 41:
1095-1098 (1997).
37. S.R Gygi, D.G. Wilkins, and D.E. Rollins. A comparison of phenobarbital and codeine incorporation into pigmented and nonpigmented rat hair. J. Pharm. Sci. 86:209-214 (1997).
38. S.J.Green and J.F. Wilson. The effect of hair color on the incorporation of methadone into hair in the rat. J. AnaL ToxicoL 20:
121-123 (1996).
39. S.P.Gygi, R.E.Joesph, E.J.Cone, D.G. Wilkins, and D.E. Rollins. Incorporation of codeine and its metabolites into hair: role of pigmentation. Drug Metab. Dispos. 24:495-499 (1996).
40. B. Gerstenberg, G. Schepers, P. Voncken, and H. VOIkeL Nicotine
and cotinine accumulation in pigmented and nonpigmented rat
hair. DrugMetab. Dispos. 23:143-148 (1995}.
41. M.H. Slawson, D.G. Wilkins, and D.E Rollins. Quantitative determination of phencydidine in pigmented and nonpigmented rat
hair by ion trap mass spectrometry. J. AnaL ToxicoL 20:350-354
(1996).
42. W.H. Harrison, R.M. Gray, and L.M. Solomon. Incorporation of damphetamine into pigmented guinea pig hair. Br. J. DermatoL 91:
415-418 (1974}.
43. R.W. Reid, F.L. O'Connor, A.G. Deakin, D.M. Ivery, and J.W.
Crayton. Cocaine and metabolites in human graying hair: pigmentary relationship. Clin. ToxicoL 34:685-690 (1996).
44. D.L. Hubbard, D.G. Wilkins, and D.E. Rollins. The incorporation
of cocaine and metabolites into hair: effects of dose an hair pigmentation. Drug Metab. Dispos. 28:1464-1469 (2000).
45. T. Uematsu, R. Sato, O. Fujimori, and M. Nakashima. Human
scalp hair as evidence of individual dosage history of haloperidol:
a possible linkage of haloperiodol excretion into hair with hair pigment. Arch. DermatoL Res. 282:120-125 (1990).
46. D.E. Rollins, D.G. Wilkins, A. Mizuno, M.H. Slawson, and C.R.
B0rges. The role of pigmentation in the distribution of drugs of
abuse in human hair. Clin. PharmacoL Thor. 67:113 (2000).
47. R. Kronstrand, S. Forstberg-Peterson, B. Kagedal, J. Ahlner, and G.
48.
49.
50.
51.
52.
53.
54.
55.
56.
57.
58.
59.
60.
61.
62.
63.
64.
65.
Larson. Codeine content in hair after oral administration of is dependent on melanin content. Clin. Chem. 45:1485-1494 (1999).
T. Mieczkowski and R. Newel. Statistical examination of hair color
as a potential biasing factor in hair analysis. Forensic Sci. Int. 107:
13-38 (2000).
C.K. Kelly, T. Mieczkowski, S. Sweeney, and J. Bourland. Hair
analysis for drugs of abuse. Hair color and race differentials or systematic differences in drug preferences? Forensic Sci. Int 107:
63-86 (2000).
T. Uematsu, N. Miyazawa, and M. Nakashima. The measurement
of ofloxacin in hair as an index of drug exposure. Eur. J. Olin.
PharmacoL 40(6): 581-584 (1991).
N. Miyazawa and T. Uematsu. Analysis of ofloxacin in hair as a
measure of hair growth and as a time marker for hair analysis. Thor.
DrugMonit. 14:525-528 (1992).
N. Nakano, T. Uematsu, H. Sato, K. Kosuge, M. Nishimoto, and M.
Nakashima. Using ofloxacin as a time marker in hair analysis for
monitoring the dosage history of haloperidol. Eur. J. Clin. PharmacoL 47:195-202 (1994).
1". Uematsu, K. Kosuge, S. Araki, M. Ishiye, Y. Asai, and M.
Nakashima. "13mecourse of appearance of ofloxacin in human
scalp hair after oral administration. ]'her. Drug Monit. 17:101-103
(1995).
N. Miyazawa, 1-. Uematsu, A. Mizuno, A. Nagashima, and M.
Nakashima. Ofloxacin in human hair determined by high performance liquid chromatography. Forensic Sci. Int. 51:65-77 (1991).
D.G. Wilkins, A. Mizuno, C.R. 8orges, and D.E. Rollins. The utility
of ofloxacin as a reference marker for hair analysis. Abstract
number 87-607, SOFT Meeting, Milwaukee, WI, October 4, 2000.
T. Uematsu, N. Miyazawa, O. Okazaki, and M. Nakashima. Possible effect of pigment on the pharmacokinetics of ofloxacin and
its excretion in hair. J. Pharm. 5ci. 81: 45-48 (1992).
A. Mizuno, T. Uematsu, and M. Nakashima. Simultaneous determination of ofloxacin, norfloxacin and ciprofloxacin in human hair
by high performance liquid chromatography and fluorescence detection. J. Chromatogr. B 653:187-193 (1994).
S. Ito and K. Wakamatsu. Chemical degradation of melanins: application to identification of dopamine-melanin. Pig. Cell Res 11:
120-126 (1998).
C.R. Borges, J. Roberts, D.G. Wilkins, and D.E. Rollins. Relationship of melanin degradation products to actuat melanin content.
Anal. Biochem. 298:116-125 (20011.
S. Ito and K. Fujita. Microanalysis of eumelanin and pheomelanin
in hair and melanomas by chemical degradation and liquid chromatography. Anal Biochem. 9:51-57 (198S).
S. Ito, K. Wakamatsu, and H. Ozeki. Spectrophotometric assayof
eumelanin in tissue samples. Anal Biochem. 215:273-277 (1993).
S. Ito. Reexamination of the structure of eumelanin. Biochem.
Biophys. Acta 883:155-161 (1986).
C. Robbins. Chemical and Physical Behavior of Human Hair, 3rd
ed. Springer Verlag, New York, NY, 1994.
E.J.Cone and R.E. Joseph. The potential for bias in hair testing for
drugs of abuse. In Drug Testing In Hair, E Kintz, Ed. CRC Press,
Boca Raton, FL, 1996, pp 69-93.
H. Chambers. Sulfonamides, trimethoprim and quinolones. In
Basic and Clinical Pharmacology, 8th ed., B. Katzung, Ed. Lange
Medical 8ooks/MacGraw Hill, New York, 2001, pp 793--802.
Manuscript received September 24, 2001;
revision received November 14, 2002.
155