e n v i r o n m e n t a l t o x i c o l o g y a n d p h a r m a c o l o g y 3 6 ( 2 0 1 3 ) 108–114

Available online at www.sciencedirect.com

j o ur nal ho me pa ge: www.elsev ier .com/ locate /e tap

Evaluation of spot urine as an alternative to 24 h urinecollection for determination of biomarkers of exposureto cigarette smoke in adult smokers�

Mohamadi Sarkar ∗, Raheema Muhammad-Kah, Qiwei Liang, Sunil Kapur,Shixia Feng, Hans RoethigAltria Client Services Inc., Center for Research and Technology, Richmond, VA 23219, USA

a r t i c l e i n f o

Article history:

Received 27 December 2012

Received in revised form

5 March 2013

Accepted 8 March 2013

Available online 15 March 2013

Keywords:

Biomarkers

Spot urine

24-h urine

Smoking

Exposure evaluation

Model

a b s t r a c t

Objective: Exposure to cigarette smoke in adult smokers (SM) can be determined by measur-

ing urinary excretion of selected smoke constituents or metabolites. Complete 24 h urine

collections are difficult to achieve in ambulatory clinical studies; therefore spot urine (SU)

might be a useful alternative. The objective of this study was to evaluate the optimum time

for SU collections, and to predict 24 h urine biomarker excretion from SU collections.

Methods: SU samples were collected at three time points (early morning, post-lunch and

evening) along with 24 h collections in 37 healthy adult smokers. Nicotine and its five

metabolites (nicotine equivalents, NE), metabolites of NNK (NNAL), pyrene (1-OHP), acrolein

(HPMA), benzene (S-PMA) and butadiene (MHBMA) were measured in 24 h and SU samples.

Correlation and agreement between creatinine-adjusted SU and 24 h urine collections were

determined from the Pearson product-moment correlation, Bland–Altman and Lin’s concor-

dance correlation analyses. A random effect regression model was used to calculate the 24 h

biomarker excretion from SU collections.

Results: There were no significant differences (p > 0.05) between the three SU collections for

the selected biomarkers of exposure except for 3-HPMA, which showed a diurnal variation.

Good correlation and statistical agreements were observed for creatinine-adjusted SU (all

three time points) and 24 h for most of the selected biomarkers. 24 h biomarker excretion

could be estimated for most of the biomarkers based on the regression model, with the early

morning SU collections giving the best results for tobacco specific biomarkers NE (R2 = 0.66)

and NNAL (R2 = 0.6).

Conclusions: SU is a useful alternative to 24 h urine collections for most of the selected

biomarkers of exposure to cigarette smoke. The early morning SU appears to be the most

feasible and practical option as an alternative to 24 h collections.

© 2013 Published by Elsevier B.V.

Abbreviations: NE, nicotine equivalents; NNK, 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone; NNAL, 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanol; 1-OHP, 1-hydroxypyrene; 3-HPMA, 3-hydroxypropylmercapturic acid; S-PMA, S-phenylmercapturic acid; MHBMA,monohydroxybutenyl mercapturic acid.

� Aspects of this paper were presented as a poster at The American Society for Clinical Pharmacology and Therapeutics (ASCPT) Annual

Meeting, Baltimore, MD, March 8–11, 2006.

∗ Corresponding author at: Health Sciences, Altria Client Services Inc.,Tel.: +1 804 335 2537.

E-mail address: mohamadi.Sarkar@altria.com (M. Sarkar).1382-6689/$ – see front matter © 2013 Published by Elsevier B.V.http://dx.doi.org/10.1016/j.etap.2013.03.001

Center for Research and Technology, Richmond, VA 23219, USA.

p h a r

1

CoplepfcseacttptHetttussirSsmuppat

i(tOeN(uageHS(m

2

TPa

e n v i r o n m e n t a l t o x i c o l o g y a n d

. Introduction

igarette mainstream smoke is a complex aerosol composedf a mixture of combustion gases and semi-volatile com-ounds, often referred to as gas/vapor phase and minute

iquid droplets representing the particulate or “tar” phase. It isstimated that there are more than 4000 different chemicalsresent in cigarette smoke (Green and Rodgman, 1996), there-ore the total systemic evaluation of human exposure to thesehemicals is almost impossible. Therefore exposure in adultmokers is often measured either by determining the urinaryxcretion of the smoke constituent or metabolites, referred tos biomarkers of exposure (Stratton et al., 2001). 24-h urineollection is considered the “gold standard” sample collec-ion technique for assessing exposure, because it estimateshe “total daily exposure” or area under the curve, which isarticularly important since the time of cigarette smokinghroughout the day is not fixed and can be highly variable.owever the 24 h urine collection method has its limitations,specially in ambulatory clinical studies. Subjects may forgeto collect samples, may lose or forget to take their collec-ion containers, or may spill the urine before collecting it inhe container. Incomplete collection of urine will result in annderestimation of the exposure to tobacco smoke in suchubjects. Spot urine (SU) collection has a good potential toerve as a satisfactory substitute to 24 h urine collection as its more convenient, less time consuming and might be moreeliable because of less frequent errors. Although the utility ofU collection has been investigated (Heavner et al., 2006), thetudy was conducted in a relatively small samples size (n = 2ales and females in each tar brand). Furthermore, the 24 h

rine collection was mathematically simulated and the timeeriod of SU collection was not systematically optimized. Theurpose of the study was to determine the feasibility of SU as

suitable alternative to 24 h urine collection and to establishhe optimum time point for SU collection.

Biomarkers of exposure evaluated in the current stud-es were: nicotine and its 5 major metabolites in urinenicotine-N-glucuronide, cotinine, cotinine-N-glucuronide,rans-3′-hydroxycotinine, and trans-3′-hydroxycotinine--glucuronide), the molar sum calculated as nicotinequivalents [NE]), total urinary NNAL (NNAL andNAL-glucuronides) for NNK [(4-(methylnitrosamino)-1-

3-pyridyl)-1-butanone)] exposure (Hecht, 1999, 2002), totalrinary 1-hydroxypyrene (1-OHP) (1-OHP, 1-OHP-glucuronidend 1-OHP-sulfate) for pyrene exposure and as a surro-ate marker for polycyclic aromatic hydrocarbons (Scherert al., 2000), urinary 3-hydroxypropylmercapturic acid (3-PMA) for acrolein exposure (Mascher et al., 2001), urinary-phenylmercapturic acid (S-PMA) for benzene exposureFeng et al., 2006), and urinary MHBMA (monohydroxy butyl

ercapturic acid) for butadiene exposure (Urban et al., 2003).

. Materials and methods

he study was performed in accordance with Good Clinicalractice and the Declaration of Helsinki criteria. The protocolsnd informed consent forms were reviewed and approved by

m a c o l o g y 3 6 ( 2 0 1 3 ) 108–114 109

the MDS Pharma Services Institutional Review Board, Lincoln,NE, before the study started. All volunteers provided writteninformed consent before enrolling in the study, were paid forparticipating, and were free to discontinue the study at anytime for any reason.

2.1. Study design

This study used a single center, controlled, open-label, obser-vational study design. Thirty seven (n = 37) healthy adult maleand female smokers who smoked conventional cigarettes(11 mg tar, 0.8 mg nicotine, 11.2 mg CO, Federal Trade Com-mission [FTC] method) participated in this study. Subjectssmoked these cigarettes during the entire duration of thestudy. Subjects were monitored for cigarette consumption onAcclimation Day to determine each subject’s maximum dailyallotment of cigarettes for the remainder of the study. Subjectswere allowed to smoke up to their daily allotment of cigarettesfrom 0700 to 2300 at predetermined times only (every 32 min)throughout the study. Controlled smoking was used to keepthe maximum number of cigarettes smoked during the studyby each subject as constant as possible. Subjects were neverforced to smoke and were allowed to smoke less than theirdaily allotment of cigarettes or even to quit smoking com-pletely at any time during the study. All urine voided by eachsubject was collected over a 24 h interval for a period of 3days (Day 1 to Day 3). The volume of urine collected duringeach interval was recorded for each subject and aliquots wereremoved for analysis of biomarkers of exposure. In addition,SU samples were taken at three time points: in the morning(first morning void upon waking, approximately 0700), postlunch (approximately 1300), and evening (approximately 1800)on all 3 days (Day 1 to Day 3). The total urine volume fromall the three spot voids were measured and recorded (includ-ing time of void) for each subject and then aliquots (Roethiget al., 2007) were removed for analysis of biomarkers of expo-sure. The remaining urine was pooled into the 24 h collectioncontainer for that subject. During the study, meals served dur-ing the confinement period were planned by a dietician tominimize dietary confounding in the biomarker assessments(Strickland et al., 1996). Urine voided in each 24 h period wasmonitored by the clinical staff. Water was permitted ad libduring the study.

3. Bioanalytical procedures

All biomarkers were measured by liquid chromatogra-phy/tandem mass spectrometry (LC-MS/MS) methods vali-dated according to the US Food and Drug AdministrationGuidance for Industry: Bioanalytical Method Validation [USDHSS FDA]. The analytical methods and validation perfor-mances for nicotine and 5 major metabolites, total NNAL,total 1-OHP, 3-HPMA S-PMA and MHBMA have been previouslyreported (Roethig et al., 2007; Sarkar et al., 2008). All analyticalbatches included appropriate calibration and quality control

samples and met the acceptance criteria according to the FDAGuidance. Urine creatinine concentration was measured usingthe rate alkaline picrate (Jaffe) method (Burtis and Ashwood,1999).

d p h

110 e n v i r o n m e n t a l t o x i c o l o g y a n

MHBMA (1,3-butadiene metabolite) was quantified inhuman urine by adding a 50-�L aliquot of a methanolic solu-tion containing the internal standard (d6-MHBMA) to a 0.5-mLsample of human urine. The sample was acidified with 1.0-mL of 10% HCl, further diluted with 0.8-mL of artificial urineand vortexed before application to a mixed mode solid phaseextraction cartridge. The Oasis HLB®, 3 cc/60 mg SPE cartridgewas previously conditioned with 2.0-mL of acetonitrile fol-lowed by 2.0-mL of 10% HCl. The sample was loaded by gravity,washed with 1.0-mL of water and eluted with 1.0-mL of 0.5%ammonium hydroxide in acetone. After evaporating the eluateto dryness at 40 ◦C, the residue was reconstituted with 200-�Lof acetonitrile prior to transfer to a silanized glass autosam-pler vial. A 40-�L aliquot was injected onto the AB/MDS SCIEXAPI 4000 LC-MS/MS system equipped with a Thermo Hyper-sil BioBasic AX, 10 × 4 mm, 5 �m pre-column and ThermoHypersil BioBasic AX, 150 × 4.6 mm, 5 �m analytical column. Acolumn switching system utilizing two mobile phases (mobilephase A: 90/10 acetonitrile/50 mM ammonium acetate pH 3.5)and (mobile phase B: 60/40 acetonitrile/50 mM ammoniumacetate pH 3.5) at a flow rate of 1.0 mL/min was used toisolate the analytes of interest into the MS/MS. The MS/MSwas operated in the negative electro-spray (ESI) mode whilemonitoring the following ion transitions for quantitation:MHBMA: m/z 232.0 → 103.2 and d6-MHBMA: m/z 238.0 → 109.1.The retention time was 8.60 min. The linear range was 0.100 to50.0 ng/mL and the R2 was >0.99. The interday precision rangedfrom 1.8 to 5.5% CV and the interday accuracy (% deviationfrom the theoretical value) ranged from −7.0% to 1.0% at fourdifferent quality control levels.

3.1. Statistical analysis

Descriptive statistics were calculated for all biomarkers(adjusted for creatinine) for the three SU collection time pointsand 24 h collections for days 1, 2 and 3. Analysis of vari-ance (ANOVA) for nested classifications was performed totest for mean differences between 24 h urine and correspond-ing SU collections at each study day. The model includedthe terms subject, study day, and urine collections nestedwithin study day. The Pearson’s product–moment correla-tion and Lin’s concordance correlation coefficient were usedto compare the relationship between the SU collections andthe corresponding 24 h urine collections in urinary levelsof selected biomarkers. The Pearson’s correlation coefficientanalysis examined the relationship between a SU collectiontaken at a particular time point on a given day with the corre-sponding 24 h urine collection for that same day. This analysisestimated precision by measuring the strength of a relationbetween two variables. Lin’s concordance correlation coeffi-cient (Lin and Torbeck, 1998) was used to evaluate the accuracyand precision between the early morning SU collections andthe 24 h urine collections. The Bland–Altman technique (Blandand Altman, 1986) was used to evaluate the closeness of agree-ment between the two collection methods. The differences

between the pooled 24 h urine collections and the pooled earlymorning SU collections (24 h − spot) were examined againstthe means of the two urine collection methods (24 h + spot)/2.The limits of agreement were obtained by calculating the

a r m a c o l o g y 3 6 ( 2 0 1 3 ) 108–114

mean difference between the two urine collection methods ± 2SD (standard deviation).

24 h biomarker excretion was calculated from spot urinedata using a random effect regression model. The modelincluded the amount of biomarker excretion in spot urineadjusted for creatinine, number of cigarette smoked per day,gender, age and body weight were used as the fixed effects andstudy day as the random effect. SAS®, PROC MIXED procedurewas used to run the model.

4. Results

A total of 37 subjects (17 men and 20 women) participatedin the study. The majority (n = 35) were Caucasians with aver-age age of 32 (11.1 SD) years, average weight of 173.7 (37.9 SD)lb, and average BMI of 26.8 (4.5 SD) kg/m2. On average, eachsubject smoked 20 cigarettes per day. The mean ± SD valuesfor creatinine (g/24 h) were not significantly different (p > 0.05)during the study (data not shown).

The mean biomarker levels collected as spot and 24 h urinecollections on Days 1–3 are presented in Table 1. The mean SUNE, total NNAL, total 1-OHP and MHBMA levels collected inthe morning, at post lunch and in the evening were not signif-icantly different (p > 0.05) between the study days. The meanSU NE, total NNAL, total 1-OHP and MHBMA levels collectedat three different times of a day and the corresponding mean24 h urine NE, total NNAL, total 1-OHP and MHBMA levels werenot significantly different (p > 0.05) for between the study days.

The mean SU total 1-OHP collected at three different timesof a day (97.35–128.44 ng/g) and corresponding mean 24 h urine1-OHP (98.45–115.06 ng/g) were not significantly different forall the study days except the post lunch and evening spotcollection on Day 1.

The mean SU 3-HPMA levels collected in the morning(477.49–619.84 �g/g) were not significantly different betweenthe study days, but were consistently lower compared to postlunch (945.51–1183.39 �g/g) and evening (1244.42–1572.53 �g/g)levels on all the 3 days. The mean SU 3-HPMA collected in themorning (477.49–619.84 �g/g) and corresponding mean 24 hurine 3-HPMA (978.67 to 1062.31 �g/g) were significantly dif-ferent (p ≤ 0.05) for Day 1, Day 2, and Day 3. There was nostatistically significant difference between the post lunch andevening SU collections and the 24 h urine collections on all thestudy days except the evening spot collection on Day 1.

The mean SU S-PMA collected in the morning(2.95–3.24 �g/g) and at post lunch (3.66 to 4.19 �g/g) werenot significantly different between the study days. Theevening SU S-PMA collection was significantly differentbetween Day 1 and Day 2. The mean SU S-PMA collected atthree different times of day (2.95–4.35 �g/g) and correspondingmean 24 h urine S-PMA (3.36–3.75 �g/g) were not significantlydifferent for all the study days.

Table 2 shows the Pearson correlation coefficient val-ues between the SU collections and 24 h urine collec-

tions on all study days; NE (r = 0.8834–0.9737), total NNAL(r = 0.7384–0.9009), total 1-OHP (r = 0.4592–0.8813), 3-HPMA(r = 0.6765–0.8270), S-PMA (r = 0.6679–0.8937), and MHBMA(r = 0.8154–0.9101).

e n v i r o n m e n t a l t o x i c o l o g y a n d p h a r m a c o l o g y 3 6 ( 2 0 1 3 ) 108–114 111

Table 1 – Creatinine-adjusted urine biomarkers of exposure in healthy adult smokers (mean ± SD).

Biomarker of Exposure Spot urine collection at 24-h urine

Early Morning (∼07:00) Post Lunch (∼13:00) Evening (∼18:00)

Nicotine Equivalents (mg/g creatinine)Day 1 11.00 ± 5.22 12.03 ± 5.62 12.65 ± 5.98 12.10 ± 5.43Day 2 10.04 ± 4.77 11.95 ± 5.70 11.83 ± 5.66 11.29 ± 5.11Day 3 11.10 ± 5.19 10.85 ± 5.63 12.71 ± 5.96 12.27 ± 6.07Total NNAL (ng/g creatinine)Day 1 223.65 ± 121.14 330.70 ± 184.47 307.15 ± 161.99 264.39 ± 145.56Day 2 225.33 ± 120.74 282.52 ± 169.17 264.87 ± 159.81 266.53 ± 149.27Day 3 233.75 ± 117.99 268.19 ± 139.78 283.70 ± 181.22 264.68 ± 144.78Total 1-OHP (ng/g creatinine)Day 1 101.33 ± 92.39 124.24 ± 50.48 128.44 ± 55.89 98.45 ± 45.71Day 2 97.35 ± 45.59 114.48 ± 48.06 115.31 ± 56.96 115.06 ± 49.69Day 3 99.41 ± 41.60 99.58 ± 48.27 122.97 ± 59.58 113.25 ± 57.453-HPMA (�g/g creatinine)Day 1 619.84 ± 344.10 1183.39 ± 640.28 1572.53 ± 705.69 978.67 ± 509.41Day 2 531.75 ± 315.45 1041.57 ± 692.96 1293.54 ± 795.44 1062.31 ± 666.71Day 3 477.49 ± 252.19 945.51 ± 670.55 1244.42 ± 671.97 1039.27 ± 648.48S-PMA (�g/g creatinine)Day 1 3.24 ± 2.32 4.19 ± 2.67 4.35 ± 2.76 3.56 ± 2.43Day 2 2.95 ± 1.92 3.82 ± 2.41 2.99 ± 2.24 3.36 ± 2.12Day 3 3.32 ± 2.34 3.66 ± 2.68 4.08 ± 3.04 3.75 ± 2.99MHBMA (�g/g creatinine)Day 1 2.32 ± 1.52 2.55 ± 1.72 3.15 ± 1.86 2.23 ± 1.59Day 2 2.22 ± 1.44 2.35 ± 1.45 2.65 ± 1.93 2.48 ± 1.65Day 3 2.25 ± 1.34 1.91 ± 1.46 2.93 ± 1.99 2.34 ± 1.38

Table 2 – Pearson correlation coefficients between creatinine-adjusted 24 h urine collections and spot urine collections.

Biomarker of exposure Pearson correlation for 24 h urine against spot urine at

Early morning (∼07:00) Post lunch (∼13:00) Evening (∼18:00)

Nicotine Equivalents (mg/g creatinine)Day 1 0.9737 0.9571 0.9430Day 2 0.9669 0.9277 0.9580Day 3 0.9404 0.8905 0.8834Total NNAL (ng/g creatinine)Day 1 0.8912 0.7867 0.9009Day 2 0.7212a 0.7958 0.9061Day 3 0.7684 0.7541a 0.7384Total 1-OHP (ng/g creatinine)Day 1 0.4592 0.8500 0.8193Day 2 0.8813a 0.8120 0.8318Day 3 0.6814 0.7009a 0.78953-HPMA (�g/g creatinine)Day 1 0.6769a 0.7903 0.8270Day 2 0.6806 0.7837a 0.7943Day 3 0.6765 0.7817a 0.7152S-PMA (�g/g creatinine)Day 1 0.8135 0.8796 0.8605Day 2 0.8937a 0.8225 0.7946Day 3 0.7682 0.7217a 0.6679MHBMA (�g/g creatinine)Day 1 0.8898 0.8905 0.8182Day 2 0.8625b 0.8592a 0.9101Day 3 0.8270 0.8154a 0.8598

Note: N = 37 samples for both 24 h urine and spot urine.a N = 36 samples for the spot urine.b N = 35 samples for the spot urine.

112 e n v i r o n m e n t a l t o x i c o l o g y a n d p h

Table 3 – Bland–Altman analysis of agreement between24 h urine excretion and spot urine measurements.

Biomarker of exposure Limits of agreement –difference (difference ± 2 SD)a

Urine nicotineequivalents (mg/g Cr)b

1.17 (−0.89, 3.23)

Total NNAL (ng/g Cr)c 38.95 (−93.73, 171.63)Total 1-OHP (ng/g Cr)d 9.03 (−59.17, 77.23)3-HPMA (�g/g Cr)d 487.83 (−263.69, 1293.34)S-PMA (�g/g Cr)c 0.41 (−1.08,1.92)MHBMA (�g/g Cr)c 0.066 (−0.82, 0.95)

a Difference in average 24 h urine excretion and spot urine ± 2 stan-dard deviation.

b Three values outside the limits.

c Two values outside the limits.d One value outside the limits.

Examination of the difference between the 24 h urine col-lections and morning SU collections (24 h − spot) against aver-age of 24 h urine and morning SU collections (24 h + spot)/2,demonstrated that greater than 95% of the data values werewithin the limits of agreement as determined by 2 standarddeviations (Table 3). The distribution of the differences aroundzero further supports the closeness of agreement betweenthe two collection methods (Table 3). Samples outside theagreement range tended to be for biomarkers with higherexcretion levels. Table 4 shows the Lin’s concordance cor-relation values between the pooled SU collections and thecorresponding 24 h urine collections for the all the selectedbiomarkers (rc = 0.6102–0.9556).

Table 5 shows the results from the random effect regressionmodel. The 24 h biomarker excretion could be estimated fromspot urine measurements. The R2 value ranged from 0.32331to 0.6903 for the regression models. Age and cigs/day did notappear to be significant factors in the regression model formost of the biomarkers. The tobacco specific biomarkers, NEand total NNAL could be estimated from the early morningspot urine measurements (R2 = 0.66 and 0.60 resp., Table 5(a)and (b)) by including weight and gender in the model.

5. Discussion

Spot urine normalized to creatinine has been frequentlyemployed to calculate exposure to tobacco biomarkers,particularly urinary nicotine and nitrosamines (Hatsukamiet al., 2004). However, no study has rigorously looked into

Table 4 – Average Lin’s concordance correlationcoefficients for creatinine-adjusted 24 h urine collectionsand morning spot urine collections for days 1–3.

Biomarker of exposure Lin’s concordancecorrelation

coefficient (rc)

Nicotine equivalents (mg/g creatinine) 0.9556Total NNAL (ng/g creatinine) 0.8197Total 1-OHP (ng/g creatinine) 0.72973-HPMA (�g/g creatinine) 0.6102S-PMA (�g/g creatinine) 0.9364MHBMA (�g/g creatinine) 0.9486

a r m a c o l o g y 3 6 ( 2 0 1 3 ) 108–114

the relationship between the two methods of urine collection(24 h and SU) for biomarkers of tobacco smoke exposure. Thisstudy was designed to evaluate the use of SU collection as analternative to 24 h urine collection for monitoring biomarkersof tobacco exposure. In order to validate the utility of theSU collections, extreme care was taken to assure completeurine collections in this confined, highly controlled researchstudy, which is difficult to achieve in ambulatory studies.The SU collection time points were chosen to be practicaland convenient, primarily to facilitate adherence to studyprotocols in ambulatory clinical studies. The results fromthe study demonstrate that the creatinine-adjusted urinarybiomarker levels are similar in all the three spot collections(early morning, afternoon and evening) and 24 h urine collec-tions for most of the selected biomarkers of tobacco exposure.Interestingly, 3-HPMA excretion appeared to be lower in themorning SU collections than in the post-lunch and evening SUcollections. This might be due to the relatively short half-lifeof this metabolite (Scherer, 2005). Even though the SU levelsfor 3-HPMA may be lower relative to the 24 h collections, thisstill allows for comparisons for differences e.g. between dayafter switching to a different product.

Generally, good correlation and accuracy was observed forthe creatinine-adjusted SU and 24 h urine collections as deter-mined by Pearson correlation coefficient and Lin’s correlationcoefficients. Pearson’s correlation coefficient is more com-monly used in method comparison. The shortcoming with thismethod is that it measures the strength of a relation betweentwo variables, not the agreement between them. It estimatesprecision but not accuracy. We observed high Lin correlationcoefficients [(rc) = 0.61–0.95] for the two methods of urine col-lection. A high degree of agreement between the two methods(greater than 95% of the samples being within limits of agree-ment) was observed through the Bland–Altman analyses.

For non-tobacco specific biomarkers the correlations werenot as strong. Possibly because these constituents are oftenconfounded by other sources of exposure, in addition tocigarette smoking exposure, like food and environmental pol-lutants (Strickland and Kang, 1999; Dor et al., 1999; Cheng et al.,2007; Tamamizu-Kato et al., 2007). The background carryoverdue to the highly variable exposure (Sarkar et al., 2008) fromother sources may have an effect on the urinary excretion lev-els when the subjects enter the clinic. The results from theregression modeling showed that early morning SU collectionscan be used to estimate 24 h biomarker excretion especially forthe tobacco specific biomarkers, NE and NNAL. This is the firstreport where a potential predictive model could be utilized toestimate 24 h urinary excretion of these biomarkers based onthe SU values, weight and gender. The R2 values >0.6 demon-strate a reasonably good correlation between the two urinecollection methods.

The validity of using the simple and straightforward crea-tinine ratio as a normalization method are corroborated byother reports (Heavner et al., 2006; Andersson et al., 2008;Carrieri et al., 2001; Mason et al., 1998) where different nor-malization techniques were compared and no significant

improvement was reported relative to creatinine adjusted val-ues. A major advantage of using the creatinine adjustmentmethod is that it allows comparison between other reportedvalues in literature (Hecht, 2002).

e n v i r o n m e n t a l t o x i c o l o g y a n d p h a r m a c o l o g y 3 6 ( 2 0 1 3 ) 108–114 113

Table 5 – Random Effect Regression Model for the prediction of 24 h urine biomarker excretion.

Slopes

Intercept Spot urine Weight Gender Cigs/day Age R2

(a) Random effect regression models for NE (mg/24 h)Post lunch (1 pm) 0.9215 0.9894 0.0472 −6.3734 NS NS 0.6503Evening (6 pm) −0.9916 0.9955 0.0542 −6.6810 NS NS 0.6940Early morning (7 am) 2.5502 1.0972 0.0382 −7.0105 NS NS 0.6603

(b) Random effect regression models for total NNAL (ng/24 h)Post lunch (1 pm) 36.6372 0.7754 1.2011 −157.58 NS NS 0.5498Evening (6 pm) −13.8956 0.8131 1.4680 −152.01 NS NS 0.5902Early morning (7 am) 46.7796 1.1678 0.9740 −165.47 NS NS 0.6001

(c) Random effect regression models for 1-OHP (ng/24 h)Post lunch (1 pm) 18.4028 0.9484 0.4050 −71.9288 NS NS 0.5325Evening (6 pm) 109.830 0.7881 NS −89.4753 NS NS 0.5935Early morning (7 am) 66.0725 0.2762 NS −73.2324 5.3084 NS 0.3400

(d) Random effect regression models for 3-HPMA (�g/24 h)Post lunch (1 pm) 327.85 0.4956 NS −572.57 45.5766 NS 0.3471Evening (6 pm) −183.39 0.3036 NS −563.56 55.3509 13.5854 0.3231Early morning (7 am) −393.71 0.8162 NS −430.49 58.9890 14.8773 0.3465

(e) Random effect regression models for S-PMA (�g/24 h)Post lunch (1 pm) −0.7037 1.0363 0.0148 −1.8541 NS NS 0.5754Evening (6pm) −0.5036 0.5298 NS −1.4224 0.2138 NS 0.3373Early morning (7 am) 0.6236 0.7884 NS −1.9726 0.1476 NS 0.3690

(f) Random effect regression models for MHBMA (�g/24 h)Post lunch (1 pm) 2.0183 0.9394 NS −1.6037 NS NS 0.6454Evening (6 pm) 0.3977 0.8001 0.0081 −1.4967 NS NS 0.6546Early morning (7 am) 1.0574 0.9863 NS −1.4957 NS 0.0278 0.6455

e = 1,

csetpei2vTtu(stSmud

2acaen

We specially thank Drs. Nancy Wang and Bruce DeGroot from

Spot urine: biomarker adjusted by urine creatinine. Gender: femalstatistically significant, p > 0.05.

These observations suggest that SU measurements, espe-ially the early morning SU (first void in the morning) isufficiently adequate for investigating relative differences inxposure to different cigarette products. This recommenda-ion serves as a practical tool, facilitating adherence to studyrotocol in large scale ambulatory studies. This considerationasily outweighs the recommendation made by others regard-ng a optimal SU time between 2 pm and 2 am (Heavner et al.,006), the robustness of which is debatable since the SU wasalidated against mathematically simulated 24 h urine values.he first morning void collection also has an added advan-

age, by overcoming any potential variability in day-time spotrine collection arising from biomarkers with short half-life

e.g. nicotine). In addition the first morning SU collection timehould provide a reasonable estimate of the exposure fromhe cigarettes smoked during the entire day. In contrast to theU collection time points reported, the urine volume of whichay be highly dependent on level of hydration, the overnight

rine collection would provide sufficient volume to measureifferent biomarkers.

SU collections may ameliorate problems associated with4 h urine collections as it is a more convenient, manage-ble and also a relatively less expensive alternative to 24 hollections. SU collections may also reduce sample integrity

nd subject-compliance issues (Carrieri et al., 2001; Arakit al., 1990). It is important to note that adjustment for creati-ine, while correcting for water dilution, introduces additional

male = 0. Cigs/day: number of cigarettes smoked per day. NS: not

variation, which must be considered when data are evalu-ated (Boeniger et al., 1993). Among the factors affecting theamount of creatinine excretion are the muscle mass of thesubject, gender, physical activity, urine flow, time of the day,diet and disease (Boeniger et al., 1993). Variability of biomark-ers can also arise from differences in how individuals smokecigarettes and how they metabolize the tobacco constituents(Sarkar et al., 2012; Liang and Sarkar, 2012). For small con-trolled clinical studies a 24 h urine collection should still beconsidered as the preferred method. However in large ambula-tory studies, cross-sectional studies or surveillance programs,where it is not feasible to obtain a reliable 24 h urine sample,the early morning spot urine samples might be considered asa useful alternate approach.

Conflict of interest

Nothing declared.

Acknowledgments

The authors thank Ms. Yan Jin for her technical assistance.

MDS Pharma Services for their statistical assistance. Clinicalconduct and bioanalysis was done at MDS Pharma Services,Lincoln, NE.

d p h

r

Determination of the major mercapturic acids of

114 e n v i r o n m e n t a l t o x i c o l o g y a n

e f e r e n c e s

Andersson, L., Haraldsson, B., Johansson, C., Barregard, L., 2008.Methodological issues on the use of urinaryalpha-1-microglobuline in epidemiological studies. Nephrol.Dial. Transplant. 23, 1252–1256.

Araki, S., Sata, F., Murata, K., 1990. Adjustment for urinary flowrate: an improved approach to biological monitoring. Int.Arch. Occup. Environ. Health 62, 471–477.

Bland, J.M., Altman, D.G., 1986. Statistical methods for assessingagreement between two methods of clinical measurement.Lancet 1, 307–310.

Boeniger, M.F., Lowry, L.K., Rosenberg, J., 1993. Interpretation ofurine results used to assess chemical exposure with emphasison creatinine adjustments: a review. Am. Ind. Hyg. Assoc. J.54, 615–627.

Burtis, C.A., Ashwood, E.R., 1999. Tietz Textbook of ClinicalChemistry. W.B. Waunders Company, Philadelphia.

Carrieri, M., Trevisan, A., Bartolucci, G.B., 2001. Adjustment toconcentration-dilution of spot urine samples: correlationbetween specific gravity and creatinine. Int. Arch. Occup.Environ. Health 74, 63–67.

Cheng, H., Sathiakumar, N., Graff, J., Matthews, R., Delzell, E.,2007. 1,3-Butadiene and leukemia among synthetic rubberindustry workers: exposure-response relationships. Chem.Biol. Interact. 166, 15–24.

Dor, F., Dab, W., Empereur-Bissonnet, P., Zmirou, D., 1999. Validityof biomarkers in environmental health studies: the case ofPAHs and benzene. Crit. Rev. Toxicol. 29, 129–168.

Feng, S., Roethig, H.J., Liang, Q., Kinser, R., Jin, Y., Scherer, G.,Urban, M., Engl, J., Riedel, K., 2006. Evaluation of urinary1-hydroxypyrene, S-phenylmercapturic acid,trans,transmuconic acid, 3-methyladenine, 3-ethyladenine,8-hydroxy-2′-deoxyguanosine and thioethers as biomarkersof exposure to cigarette smoke. Biomarkers 11, 28–52.

Green, C.R., Rodgman, A., 1996. The Tobacco Chemists’ ResearchConference: a half century forum for advances in analyticalmethodology of tobacco and its products. Recent Adv. Tob. Sci.22, 131–304.

Hatsukami, D.K., Lemmonds, C., Zhang, Y., Murphy, S.E., Le, C.,Carmella, S.G., Hecht, S.S., 2004. Evaluation of carcinogenexposure in people who used “reduced exposure” tobaccoproducts. J. Natl. Cancer Inst. 96, 844–852.

Heavner, D.L., Morgan, W.T., Sears, S.B., Richardson, J.D., Byrd,G.D., Ogden, M.W., 2006. Effect of creatinine and specificgravity normalization techniques on xenobiotic biomarkers insmokers’ spot and 24-h urines. J. Pharm. Biomed. Anal. 40,928–942.

Hecht, S.S., 1999. Tobacco smoke carcinogens and lung cancer. J.

Natl. Cancer Inst. 91, 1194–1210.

Hecht, S.S., 2002. Human urinary carcinogen metabolites:biomarkers for investigating tobacco and cancer.Carcinogenesis 23, 907–922.

a r m a c o l o g y 3 6 ( 2 0 1 3 ) 108–114

Liang, Q., Sarkar, M., 2012. Intra- and inter-individual variabilityin urinary nicotine excretion and plasma cotinine in adultcigarette smokers. Regul. Toxicol. Pharmacol. 64, 388–393.

Lin, L., Torbeck, L.D., 1998. Coefficient of accuracy andconcordance correlation coefficient: new statistics formethods comparison. PDA J. Pharm. Sci. Technol. 52, 55–59.

Mascher, D.G., Mascher, H.J., Scherer, G., Schmid, E.R., 2001.High-performance liquid chromatographic-tandem massspectrometric determination of 3-hydroxypropylmercapturicacid in human urine. J. Chromatogr. B: Biomed. Sci. Appl. 750,163–169.

Mason, H.J., Williams, N.R., Morgan, M.G., Stevenson, A.J.,Armitage, S., 1998. Influence of biological and analyticalvariation on urine measurements for monitoring exposure tocadmium. Occup. Environ. Med. 55, 132–137.

Roethig, H.J., Zedler, B.K., Kinser, R.D., Feng, S., Nelson, B.L., Liang,Q., 2007. Short-term clinical exposure evaluation of asecond-generation electrically heated cigarette smokingsystem. J. Clin. Pharmacol. 47, 518–530.

Sarkar, M., Kapur, S., Frost-Pineda, K., Feng, S., Wang, J., Liang, Q.,Roethig, H.J., 2008. Evaluation of biomarkers of exposure toselected cigarette smoke constituents in adult smokersswitched to carbon filtered cigarettes in short term and longterm clinical studies. Nicotine Tob. Res., 1–12.

Sarkar, M., Wang, J., Liang, Q., 2012. Metabolism of nicotine and4-(methylnitrosamino)-l-(3-pyridyl)-l-butanone (NNK) inmenthol and non-menthol cigarette smokers. Drug Metab.Lett..

Scherer, G., 2005. Biomonitoring of inhaled complex mixtures –ambient air, diesel exhaust and cigarette smoke. Exp. Toxicol.Pathol. 57 (Suppl. 57), 75–110.

Scherer, G., Frank, S., Riedel, K., Meger-Kossien, I., Renner, T.,2000. Biomonitoring of exposure to polycyclic aromatichydrocarbons of nonoccupationally exposed persons. CancerEpidemiol. Biomarkers Prev. 9, 373–380.

Stratton, K., Shetty, P., Wallace, R., Bondurant, S., 2001. Clearingthe smoke: the science base for tobacco harm reduction –executive summary. Tob. Control 10, 189–195.

Strickland, P., Kang, D., 1999. Urinary 1-hydroxypyrene and otherPAH metabolites as biomarkers of exposure to environmentalPAH in air particulate matter. Toxicol. Lett. 108, 191–199.

Strickland, P., Kang, D., Sithisarankul, P., 1996. Polycyclic aromatichydrocarbon metabolites in urine as biomarkers of exposureand effect. Environ. Health Perspect. 104 (Suppl. 5), 927–932.

Tamamizu-Kato, S., Wong, J.Y., Jairam, V., Uchida, K., Raussens,V., Kato, H., Ruysschaert, J.M., Narayanaswami, V., 2007.Modification by acrolein, a component of tobacco smoke andage-related oxidative stress, mediates functional impairmentof human apolipoprotein E. Biochemistry 46, 8392–8400.

Urban, M., Gilch, G., Schepers, G., van Miert, E., Scherer, G., 2003.

1,3-butadiene in human and rat urine using liquidchromatography with tandem mass spectrometry. J.Chromatogr. B: Analyt. Technol. Biomed. Life Sci. 796, 131–140.