Normalization of Urinary Drug Concentrations with Specific Gravity and Creatinine

Journal of Analytical Toxicology, Vol. 33, January/February 2009 Normalization of Urinary Drug Concentrations with Specific Gravity and Creatinine Ed...
Author: Dana Johnston
1 downloads 0 Views 128KB Size
Journal of Analytical Toxicology, Vol. 33, January/February 2009

Normalization of Urinary Drug Concentrations with Specific Gravity and Creatinine Edward J. Cone1,*, Yale H. Caplan2, Frank Moser3, Tim Robert3, Melinda K. Shelby3, and David L. Black3 1Johns

Hopkins School of Medicine, Department of Psychiatry and Behavioral Sciences, Baltimore, Maryland 21224; Scientific Services, 3411 Phillips Drive, Baltimore, Maryland 21208; and 3Aegis Sciences Corporation, 515 Great Circle Road, Nashville, Tennessee 37228 2National

Abstract Excessive fluid intake can substantially dilute urinary drug concentrations and result in false-negative reports for drug users. Methods for correction (“normalization”) of drug/metabolite concentrations in urine have been utilized by anti-doping laboratories, pain monitoring programs, and in environmental monitoring programs to compensate for excessive hydration, but such procedures have not been used routinely in workplace, legal, and treatment settings. We evaluated two drug normalization procedures based on specific gravity and creatinine. These corrections were applied to urine specimens collected from three distinct groups (pain patients, heroin users, and marijuana/ cocaine users). Each group was unique in characteristics, study design, and dosing conditions. The results of the two normalization procedures were highly correlated (r = 0.94; range, 0.78–0.99). Increases in percent positives by specific gravity and creatinine normalization were small (0.3% and –1.0%, respectively) for heroin users (normally hydrated subjects), modest (4.2–9.8%) for pain patients (unknown hydration state), and substantial (2- to 38-fold increases) for marijuana/cocaine users (excessively hydrated subjects). Despite some limitations, these normalization procedures provide alternative means of dealing with highly dilute, dilute, and concentrated urine specimens. Drug/metabolite concentration normalization by these procedures is recommended for urine testing programs, especially as a means of coping with dilute specimens.

Introduction Urination is one of the primary methods for elimination of xenobiotics and their respective metabolites from the body. The peak urinary elimination period for drugs and metabolites in humans generally occurs over the first 24 h following exposure. In the absence of additional exposure, concentrations gradually diminish, thereafter, over a period of days to weeks depending upon the physico-chemical nature of the drug * Author to whom correspondence should be addressed: Edward J. Cone, 441 Fairtree Drive, Severna Park, MD 21146. E-mail: [email protected].

species. Over the entire period of elimination, urinary concentrations can vary widely and correlate inversely with urinary flow (1). The volume of fluid intake during the period of elimination is an important factor in determining urine flow and urinary drug concentration. Ingestion of excess fluids results in dilution, whereas fluid restriction produces concentration of urinary components. Individual control over fluid intake and its consequential influence on urinary concentrations has become well recognized among drug abusers. A variety of commercial sites advertise that use of their product allows one to “beat a drug test”. Although many claims are made for these products, the primary basis for most products is their emphasis upon ingestion of large volumes of fluid. Definitive drug dosing studies have also demonstrated the effectiveness of fluid ingestion on reduction of drug concentrations below administrative cutoff concentrations (2). Methods utilized in drug testing programs to minimize the effect of excessive dilution frequently include measuring specific gravity and creatinine. “Dilute” urine specimens are defined by the United States Department of Health and Human Services (DHHS) drug testing guidelines as those specimens whose specific gravity is > 1.0010, but < 1.0030 and their creatinine concentration is ≥ 2.0, but < 20.0 mg/dL (3). However, despite identification of some specimens as dilute, no further testing is allowed unless drug/metabolite concentration exceeds the administrative cutoff concentration. In the fields of sports testing (4–6) and environmental monitoring for toxins (1,7–9), somewhat different approaches are utilized to account for fluctuating urinary concentrations. Correction of urine concentrations (“normalization” procedures) is frequently based on specific gravity or creatinine measures. Normalization of urine concentrations based on specific gravity measurements are generally based on the Levine-Fahy equation (10), which adjusts concentrations to a reference value appropriate for the population. A similar correction, based on creatinine, can also be used to account for fluctuating urine concentrations. In this study, we sought to determine if specific gravity and creatinine normalization of drug/metabolite concentrations

Reproduction (photocopying) of editorial content of this journal is prohibited without publisher’s permission.

1

Journal of Analytical Toxicology, Vol. 33, January/February 2009

would be useful to account for the variation in urine flow and subsequent dilution of drug analytes in urine. To account for the large variation that may occur in individual excretion patterns and daily fluid consumption, corrections were performed with urine specimen data from three different populations who were administered or were using licit and/or illicit drugs.

Methods Specimens

Urine specimens in this study originated from three different populations: pain patients; heroin users; and cannabis/cocaine users. The pain population consisted of patients chronically treated with opioids and/or other therapeutic agents who provided informed consent to participate in drug monitoring programs. A total of 10,922 urine specimens obtained over the period of January through December 2006 from 31 pain clinics located in 6 states (TN, WV, KY, OH, FL, IN) were collected and shipped to Aegis Sciences (Nashville, TN) for analysis of prescribed and illicit drugs. Patients’ state of hydration at the time of collection was not recorded. A total of 10,899 specimens had concurrent specific gravity and creatinine measures and drug/metabolite concentration data available for evaluation. The heroin users were six healthy male subjects who provided informed consent to participate in a controlled heroin dosing study. The study was approved by an institutional review board. The subjects resided on the clinical ward of the National Institute of Drug Abuse (NIDA) (Baltimore, MD) throughout the duration of the study. During dosing sessions, subjects were administered single doses of heroin hydrochloride by the intranasal route (6 mg or 12 mg) and by the intramuscular route (6 mg). Subjects were allowed to drink fluids ad libitum throughout the study. All urine specimens were collected throughout the study. The details of the urine excretion study have been published (11). The cannabis/cocaine users were four healthy males and one female subject who provided informed consent to participate in a controlled marijuana and cocaine dosing study. The study was approved by an institutional review board. The subjects resided on the clinical ward of NIDA (Baltimore, MD) throughout the duration of the study. During dosing sessions, subjects were administered a single dose of marijuana [3.58% ∆9-tetrahydrocannabinol, (THC)] by the smoked route on day 1 of the study. On day 2, approximately 22 h after drug administration, subjects drank specified quantities of fluids. One of the fluid intake regimens involved drinking 1-gal of fluid divided into 1-qt aliquots in hourly intervals. This study sequence was repeated on days 3 and 4, but subjects received an intranasal dose of cocaine hydrochloride (40 mg) on day 3. All urine specimens were collected throughout the study. The details of the urine excretion study have been published (2). Specific gravity, creatinine, immunoassay, and confirmation assays

All assays of pain patient specimens were performed at Aegis Sciences. Specific gravity measurements of urine specimens

2

from pain patients were performed with a Rudolph Research Analytical J57 Automatic Refractometer (Flanders, NJ). All specific gravity data were recorded to four decimal places. The response for control samples containing sodium chloride targeted for a range of 1.0020 to 1.0399 was linear (r = 0.999, N = 10 control samples, 5 runs). Mean responses across the range varied from 0.1% to 3.0% of the target value. Urinary creatinine measurement of urine specimens was performed with DRI® Creatinine-Detect Test® reagents (Microgenics, Fremont, CA) on a Hitachi 717 automated analyzer (Boehringer Mannheim/ Hitachi, Indianapolis, IN). The manufacturer’s stated limit of linearity is from 0.78 to 420 mg/dL. In practice, the response for control samples containing creatinine, targeted for a range of 1.0 to 500 mg/dL, was linear from 1.0 to 125 mg/dL (r = 0.999, N = 11 control samples, 5 runs) but appeared non-linear from 250 to 500 mg/dL. Responses across the range of 1.0 to 125 mg/dL varied from 1.0% to 12% of the target value. For the 250 and 500 mg/dL control samples, the responses were lower than the target values by 16.5% and 37.3%, respectively. All urine specimens from pain patients were analyzed for licit and illicit drugs at Aegis Sciences. The specimens were screened by immunoassay and confirmed by gas chromatography–mass spectrometry (GC–MS) as previously described (12). All urine specimens from the heroin and marijuana/cocaine subjects were analyzed for creatinine and specific gravity at NIDA. In addition, the specimens were screened by immunoassay and confirmed for drugs and/or metabolites by GC–MS as previously described (2,11). Specific gravity and creatinine normalization

The procedure for specific gravity normalization of drug/ metabolite concentrations in urine was based on the LevineFahy equation (10) as follows: ConcentrationSG normalized = Concentrationspecimen • (SGreference – 1)/(SGspecimen – 1)

Where drug/metabolite concentration is expressed in nanograms per milliliter, SGreference is a population reference value for specific gravity representing “normal” or “nondiluted urine” and SGspecimen is the specific gravity of the test specimen. The SGreference utilized in all corrections was 1.0200 (13). The procedure developed in this study for creatinine normalization of drug/metabolite concentrations in urine was based on the following equation: ConcentrationCR normalized = Concentrationspecimen • (CRreference)/(CRspecimen)

Where drug/metabolite concentration is expressed in nanograms per milliliter, CRreference is a population reference concentration (mg/dL) for creatinine representing “nondiluted urine” and CRspecimen is the creatinine concentration (mg/dL) of the test specimen. The CRreference utilized in all corrections in this study was 100 mg/dL; a value similar to the derived mean creatinine concentrations observed in the pain patient population reported in this study. It should be noted that the

Journal of Analytical Toxicology, Vol. 33, January/February 2009

procedure developed in this study for creatinine normalization differs from other procedures, also referred to as “creatinine normalization”, which involves calculation of the ratio of drug/metabolite concentration to creatinine concentration (14). Specimens obtained from the pain patient population were rejected from normalization if their specific gravity measures were < 1.0020 and/or creatinine measures were < 10 mg/dL. No limits on specific gravity or creatinine were used for heroin users. Specimens from the cannabis/cocaine users were rejected from normalization if their specific gravity measures were less than 1.0010.

Results Correlation of urinary specific gravity and creatinine in pain patients

Histograms of specific gravity and creatinine values from 10,899 pain patient urine specimens are illustrated in Figure 1. The average, median, standard deviation, and range were as follows: 1.0166, 1.0164, 0.0085, and 1.0000–1.0549, respectively (specific gravity), and 100.9 mg/dL, 94.1 mg/dL, 57.9 mg/dL, and 0–300.1 mg/dL, respectively (creatinine). A total of 17 (0.16%) specimens had specific gravity measures equal to 1.0000. These 17 specimens contained creatinine in the range of 14.9–57.0 mg/dL. Six (0.06%) specimens had specific gravity measures in the range of > 1.0000–1.0009, and their corresponding creatinine measures were in the range of 2.0–52.3 mg/dL. A total of 164 (1.5%) specimens had specific gravity measures in the range of > 1.0010 to < 1.0030; 128 (78.0%) of these specimens met DHHS criteria (3) as “dilute” (specific

mg/dL

Figure 1. Histograms of specific gravity (A) and creatinine (B) for 10,899 pain patients.

gravity > 1.0010 to < 1.0030 and creatinine ≥ 2.0 to < 20.0 mg/dL). A total of 6969 (63.9%) specimens had specific gravity measures < 1.0200. The majority of specimens had specific gravity in the range 1.0030–1.0400 (N = 10658, 97.8%). At the upper level of specific gravity measures, 54 (0.5%) specimens had specific gravity measures > 1.0400. A total of five urine specimens (0.05%) had creatinine measures < 2.0 mg/dL (range 0–1.1 mg/dL). The specific gravity measures for these specimens were in the range of 1.0040 to 1.0178; thus, none met current DHHS criteria (creatinine < 2.0 mg/dL and specific gravity ≤ 1.0010 or ≥ 1.0200) as “substituted”. There were 419 specimens (3.8%) with creatinine measures in the range of ≥ 2.0 to < 20.0 mg/dL. As noted, only 128 (30.5%) of these had matching specific gravity results consistent with criteria by DHHS to qualify as dilute specimens. A total of 5831 (53.5%) specimens had creatinine measures 200.0 mg/dL. A linear correlation of specific gravity and creatinine for the entire data set of paired urine specimens (N = 10,889) is shown in Figure 2. The correlation coefficient by linear regression was 0.84. Normalization of urinary drug concentrations in pain patients

Drug/metabolite concentrations for 28 analytes measured by GC–MS in urine specimens from pain patients were normalized by the specific gravity and creatinine procedures. Correlations of corrected concentrations by these procedures were generally quite high as shown in Table I. The mean correlation coefficient (r) across all analytes was 0.88 with a range of 0.61–0.98. The influence of normalization on positivity rate is summarized in Table II. To determine the widest possible effect of these correction procedures, the uncorrected median concentration for each drug/metabolite was utilized as the cutoff concentration. In this evaluation, if a specimen with an uncorrected concentration below the median was corrected to a concentration greater than the median, this resulted in a change from “negative” to “positive” and vice versa for specimens above the median. For example, the median concentra-

Figure 2. Correlation of urinary specific gravity and creatinine in 10,899 pain patient specimens. Linear regression of the data (grey line) produced a correlation coefficient (r) of 0.84.

3

Journal of Analytical Toxicology, Vol. 33, January/February 2009

tion for amphetamine was 3910 ng/mL. If a specimen had an uncorrected concentration of 3534 ng/mL and a specific gravity corrected concentration of 5013 ng/mL, this represented a change from negative to positive. Alternately, if an amphetamine positive specimen had an uncorrected concentration of 4472 ng/mL and a specific gravity corrected concentration of 3494 ng/mL, this represented a change from positive to negative. The same type of evaluation was performed based on the creatinine correction procedure. The overall effects of these correction procedures were summarized by subtraction of the percent “new negatives” from “new positives” to provide assessment of the overall effect. With the exception of phentermine, more new positives than new negatives were generated by both normalization procedures. For codeine, specific gravity normalization produced more positives, whereas creatinine normalization produced fewer positives. Specific gravity and creatinine normalizations produced an overall average increase in “new positives” of 9.8% and 4.2%, respectively. Using a paired t-test, these data for

the two types of normalization procedures were significantly different (p < 0.001). Hence, specific gravity normalization produced an approximate twofold overall increase in “new positives” relative to creatinine. Normalization of urinary morphine concentrations of heroin users

The effect of specific gravity and creatinine normalizations for urine specimens from six normally hydrated heroin subjects (N = 313 specimens) was relatively small. Overall creatinine normalizations for morphine tended to be slightly lower than specific gravity normalizations. Mean corrected concentrations by both procedures were significantly lower (specific gravity, p < 0.05; creatinine, p < 0.01) compared to uncorrected concentrations. Figure 3 illustrates corrected and uncorrected urinary concentrations of morphine for a heroin user who was administered 6 mg of heroin HCl intramuscularly under controlled conditions. The effects of these normalizations on the production of “new positives” and “new negatives” at cutoff concentrations of 300 and 2000 ng/mL were small and mixed in direction. The Table I. Correlation of Specific Gravity and Creatinine Normalized Drug overall changes in positivity by specific gravity Data for Pain Patient Specimens correction were 0.3% and –1.0%, respectively, at the two cutoff concentrations. The changes in # Regression Correlation Drug Specimens Equation* Coefficient (r) positivity by creatinine normalizations for these specimens were –1.0% and –1.3%, respectively.

α-HO-alprazolam Amphetamine Benzoylecgonine Butalbital Carisoprodol Clonazepam Codeine Desmethyldiazepam Dihydrocodeine EDDP† Fentanyl Hydrocodone Hydromorphone Lorazepam Meperidine Meprobamate Methadone Methamphetamine Morphine Normeperidine Norpropoxyphene Oxazepam Oxycodone Oxymorphone Phentermine Propoxyphene Temazepam THCCOOH

776 121 303 297 270 159 135 853 2278 1118 455 5717 3684 83 35 565 1042 41 1050 40 344 1513 2056 1622 34 191 1276 958

y = 0.6662x + 91.555 y = 1.0006x – 961.51 y = 0.6884x + 1414.1 y = 1.1709x – 364.55 y = 0.7039x + 1705.6 y = 0.5112x + 163.8 y = 1.0163x – 864.29 y = 0.7401x + 48.051 y = 0.8828x + 4.7504 y = 0.6716x + 859.1 y = 0.5224x + 25.08 y = 1.0206x – 331.41 y = 0.7772x + 107.73 y = 0.6567x + 239.24 y = 0.7354x + 733.56 y = 1.2506x – 9764.7 y = 0.6884x + 578.17 y = 0.8345x – 537.96 y = 0.8925x – 1227.4 y = 0.7694x + 542.09 y = 0.9386x + 216.41 y = 0.7793x + 48.616 y = 0.7935x – 2.2875 y = 0.6961x + 498.41 y = 0.5817x + 6330.5 y = 1.1461x – 228.28 y = 0.6157x + 457.56 y = 0.6635x + 17.518

Mean Range * x = specific gravity and y = creatinine. † Abbreviations: THCCOOH, 11-nor-9-carboxy-∆9-tetrahydrocannabinol and EDDP, 2-ethylidene-1,5-dimethyl-3,3-diphenylpyrrolidine.

4

0.979 0.954 0.957 0.940 0.937 0.779 0.956 0.904 0.962 0.935 0.795 0.954 0.965 0.928 0.970 0.945 0.919 0.989 0.949 0.977 0.927 0.937 0.973 0.950 0.907 0.889 0.946 0.962 0.935 0.779–0.989

Normalization of urinary drug metabolite concentrations of marijuana/cocaine users following excessive fluid intake

The effect of specific gravity and creatinine normalizations were substantial with urine specimens from the five subjects who participated in the excessive hydration study (drinking 1 gal of fluids) following marijuana and cocaine administration (2). Average baseline measures of specific gravity and creatinine for these subjects prior to excessive hydration were 1.0156 (range = 1.0080–1.0240) and 116.0 (range = 76.0–204.0), respectively. On day 1 of the study, subjects smoked a single marijuana cigarette (3.58% THC), and on day 3, they received a 40-mg dose of cocaine hydrochloride by the intranasal route. Ingestion of 1 gal of fluids (divided into 1-qt aliquots administered hourly) on days 2 and 4 produced highly dilute urine specimens with specific gravities < 1.0030 and creatinines < 20 mg/dL starting approximately 1.5 to 2.0 h following commencement of drinking. These measures returned to baseline values within 4 to 6 h following drinking. Marijuana and cocaine concentrations dropped rapidly below cutoff concentrations during and following excessive drinking. By the time subjects had ingested 2 qt of fluid, they were generally producing specimens that tested negative for drug metabolites. Recovery of uncorrected urine drug metabolite concentrations to pre-

Journal of Analytical Toxicology, Vol. 33, January/February 2009

dilution concentrations occurred over a period of 8–10 h. Cocaine metabolite concentrations frequently returned to positive after excess fluid was eliminated, but marijuana metabolite concentrations rarely returned to positive. Specific gravity and creatinine concentration normalizations for specimens collected during and after excessive fluid ingestion reverted many of the excessively dilute negative specimens to positives. A summary of the uncorrected and corrected data for specimens during and following the period of ingestion of 1-gallon of fluid is shown in Table III. Specific gravity and creatinine normalizations produced from 3- to 38-fold increases in percent positives for marijuana metabolite and 2- to 3-fold increases for cocaine. The effect of concentration corrections on the excretion profile of one subject during excessive hydration is illustrated in Figure 4.

Discussion

Urinary drug and metabolite concentrations are known to be highly influenced by individual patterns of fluid intake on a daily basis. This study demonstrated that concentration adjustments (normalization) by specific gravity or creatinine take into account a large portion of the variation that occurs from different populations of drug users with different patterns of fluid intake. The use of these normalization procedures could potentially eliminate much of the control exerted by a drug user over the outcome of a drug test by attempted dilution prior to the test. In this study, both specific gravity and creatinine normalizations were evaluated in different types of test conditions and drug-using populations with the goal of determining if adjustments by both procedures were comparable and the extent of correction effects on overall positivity rates. Specific gravity normalizations of drug concentrations in specimens from pain Table II. Effect of Specific Gravity and Creatinine Normalization Procedures on patients produced an approximate twofold Overall Change in % Positives for Pain Patient Specimens* increase in %positives compared to creatinine normalizations when evaluated at Median % Positive % Positive the median concentration of each analyte Concentration Change Change as the cutoff concentration (Table II). The # (Cutoff) by SG† by CR same normalization procedures, when Drug/Metabolite Specimens (ng/mL) Correction Correction applied to “normally hydrated” heroin users at cutoff concentrations of 300 and α-HO-Alprazolam 766 296 26.6 14.8 2000 ng/mL, produced small or negligible Amphetamine 121 3910 9.1 2.5 changes in the overall %percent positive Benzoylecgonine 303 1069 3.0 3.6 Butalbital 297 829 12.8 8.4 rates. However, when normalization proCarisoprodol 270 2504.5 12.2 7.0 cedures were applied to specimens from Clonazepam 159 380 14.5 0.6 subjects who drank excessive fluid (1 gal Codeine 135 2857 6.7 –5.2 over 4 h) following marijuana (20/50 Desmethydiazepam 853 305 16.6 7.9 ng/mL cutoff concentration) and cocaine Dihydrocodeine 2278 273 12.4 5.0 use 150/300 ng/mL cutoff concentration), EDDP 1118 3421 11.5 5.8 a substantial increase in %positives was Fentanyl 455 23 9.2 7.9 observed (Table III). Consequently, it Hydrocodone 5717 1387 14.8 9.6 appears that these normalization proceHydromorphone 3684 477 12.3 5.5 dures were most effective for dilute and Lorazepam 83 824 9.6 2.4 highly dilute specimens. Meperidine 35 1437 5.7 8.6 Meprobamate 565 13,450 9.6 7.1 The availability of the large data-set of Methadone 1042 2203.5 10.3 4.9 urine specimens from the pain populaMethamphetamine 41 1854 4.9 0.0 tion also allowed evaluation of the relaMorphine 1050 10,502 7.1 6.0 tionship of specific gravity to creatinine. Normeperidine 40 1138 10.0 2.5 Specific gravity and creatinine measures Norpropoxyphene 344 4486 11.3 7.6 generally have been reported to be highly Oxazepam 1513 690 14.4 9.8 correlated both for individual specimens Oxycodone 2056 2716.5 10.2 5.7 (“spot samples”) (5,7,15,16) and 24-h colOxymorphone 1622 1639 8.8 4.4 lections (15). A similar high correlation (r Phentermine 34 41,983 –17.6 –32.4 = 0.84) was observed for specific gravity Propoxyphene 191 558 6.8 6.3 and creatinine measures in the 10899 Temazepam 1276 635 13.6 7.8 THCCOOH 958 41 7.8 3.9 specimens obtained from the pain population in this study (Figure 2). CorrelaMean 9.8 4.2 tions of corrected drug concentrations SD 7.0 8.1 (Table I) for these patients also were high across the 28 drug/metabolite groups. To * The median drug concentration was used as the cutoff concentration to determine if normalized specimens were positive or negative. Percent positive change was calculated by subtraction of % negatives from % positives. the authors’ knowledge, this is the first † Abbreviations: SG, specific gravity; CR, creatinine; THCCOOH, 11-nor-9-carboxy-∆9-tetrahydrocannabinol; EDDP, comparison of specific gravity/creatinine 2-ethylidene-1,5-dimethyl-3,3-diphenylpyrrolidine; and SD, standard deviation. relationships and corresponding drug

5

Journal of Analytical Toxicology, Vol. 33, January/February 2009

concentration adjustments across a large date-set of confirmed creatinine in highly concentrated specimens would produce drug results. overestimation of drug/metabolite concentrations by the creIt is important to note that both specific gravity and creatiatinine correction procedure. nine normalization procedures employed in this study involved Other limitations should also be recognized in the use of speadjustments based on population reference values of specific cific gravity and creatinine for urinary concentration adjustgravity and creatinine. The reference values for specific gravity ments. It is known that a number of physiological factors can and creatinine utilized in this study were 1.0200 and 100 influence specific gravity and creatinine measures. Urine spemg/dL, respectively. Other investigators have used specific cific gravity measures reflect the amount of dissolved subgravity reference values in the range of 1.018 to 1.024 for corstances present. Glucose, protein, or dyes used in diagnostic rection (1,8,10,17). A reference value of 1.0200 is used by the tests increase specific gravity, and certain disease states (e.g., World Anti-Doping Agency (WADA) for normalization of testosdiabetes insipidus, glomerulonephritis, renal failure) can proterone precursors, metabolites, and other endogenous steroids duce lower than expected specific gravity measures. Creatiin urine (6). Almost all anti-doping programs worldwide adhere nine concentration in urine is increased by exercise and the to the WADA procedures for normalization of endogenous presence of cooked meat in the diet and is age-, weight-, and steroids. In addition, some environmental and sports procesex-dependent (7,18). Women have significantly lower values of dures reject specimens from normalization that are considered creatinine and specific gravity than men, and older subjects (> either “too dilute” or “too concentrated” by placing defined 50 y) have lower creatinine than younger subjects (7). limits on specific gravity and creatinine. For example, Allession Despite limitations associated with specific gravity and creet al. (9) suggested rejection of specimens that had specific atinine urinary measures, the normalization procedures utigravity measures that were < 1.010 or > 1.030 and specimens lized in this study generally improved drug detection and poswith creatinine < 50 mg/dL or > 300 mg/dL. itivity rates, particularly with troublesome “dilute” urine For the purposes of this study, urine specimens from the specimens that contained drug/metabolite(s) below adminispain population data-set were rejected for normalization if tration cutoff concentrations. Concentration normalization of their specific gravity measures were < 1.0020 and/or creatinine these highly dilute specimens converted many of the “dilute measures were < 10 mg/dL. These limits restricted concentration adjustment for this population to a maximum of 10-fold correction. No limits were necessary for the data-set from heroin users as their specimens were generally within normal limits for specific gravity and creatinine. These subjects were allowed to drink fluids ad libitum during the course of the study and were considered “normally hydrated”. Limits were placed on the marijuana/cocaine users because the study design involving drinking excessive volumes of fluids. Normalization of specimens from this group of drug users were not made if the specific gravity measure was < 1.0010. In general, the use of limits on specific gravity and creatinine should take into account issues such as the nature of the population being tested, analytical accuracy of specific gravity and creatinine Figure 3. Example of specific gravity (SG) and creatinine (CR) normalizaassays, and programmatic needs. tions of morphine concentrations excreted in urine by a normally hydrated Based on control data generated in the current study, it heroin user. Specimens were collected following administration of 6 mg appears that the linearity and accuracy of the specific gravity heroin hydrochloride by the intramuscular route and analyzed by GC–MS. assay extended over a broader range than the creatinine assay. Measures of specific gravity of control samples across a range 1.0020 to 1.0399 Table III. Effect of Specific Gravity and Creatinine Normalizations of Excessively varied from 0.1% to 3.0% of target Dilute Urine Specimens* values. Creatinine measures of control Cutoff % Positive % Positive % Positive samples across a range of 1 to 125 mg/dL Concentration Without by SG by CR varied by 1% to 12%. At higher creatinine # (ng/mL) Correction Correction Correction concentrations, the assay underestimated target concentrations of control samples. Marijuana 71 50 1.4 49.3 53.5 For example, samples containing 250 and Marijuana 71 20 22.5 90.1 74.6 500 mg/dL creatinine were underestiCocaine 75/74 † 300 25.3 86.5 84.0 Cocaine 75/74 † 150 46.7 95.9 96.0 mated by 16.5% and 37.3%, respectively. Thus, it is critical that those involved in * The data represent changes in normalized urine concentrations of THCCOOH or benzoylecgonine for five use of specific gravity and creatinine subjects who smoked one marijuana cigarette (3.58% THC) or received intranasal cocaine hydrochloride (40 mg). On the day following drug administration, each subject drank 1 gal of fluid over a period of 4 h. Corrections were adjustments to concentration understand applied only to those specimens collected following the start of fluid intake. the analytical characteristics of their † One specimen had a specific gravity of 1.0000 and could not be corrected. assays. The effect of underestimation of

6

Journal of Analytical Toxicology, Vol. 33, January/February 2009

negatives” to “positives”. It also should be noted that, in some cases, highly concentrated specimens became “normalized” to lower concentrations. How to deal with concentrated specimens has rarely been addressed in urine testing programs except in special situations where an individual is being monitored over time for new or continued drug use. Concentration corrections in these situations could provide a more systematic means of differentiating new use from prolonged excretion resulting from earlier drug use. Although simplistic in concept, it is also worth stating that normalized data must be compared to a cutoff concentration that represents a “normalized” cutoff concentration. For example, the DHHS workplace confirmation cutoff concentration for benzoylecgonine (cocaine metabolite) is 150 ng/mL. This cutoff concentration is utilized to determine if a specimen is positive or negative regardless of whether a given specimen is highly dilute or highly concentrated. Instead, if drug/metabolite concentration data are “normalized”, then individuals with highly dilute urine specimens may be identified as “positive” if they exceed the “normalized” cutoff concentration and individuals with highly concentrated urine specimens, after correction, may be negative. Alternately, corrections could be applied only to “dilute” specimens. Overall, it would seem that the use of specific gravity and/or creatinine normalizations procedures in association with “normalized” cutoff concentrations is worthy of consideration and could enhance the accuracy and credibility of drug testing and monitoring programs. It is recommended that these normalization procedures receive careful consideration in urine testing programs especially as a means of coping with dilute specimens.

Figure 4. Examples of specific gravity (SG) and creatinine (CR) normalizations of cocaine (A) and marijuana (MJ) (B) metabolites excreted in urine after deliberate dilution (24 h after drug administration) by drinking 4 qt of fluid. The dotted line illustrates cutoff concentrations for cocaine metabolite (300 ng/mL) and marijuana metabolite (50 ng/mL). The arrows indicate the times of ingestion of the 4 qt of fluid.

References 1. S. Araki, F. Sata, and K. Murata. Adjustment for urinary flow rate: an improved approach to biological monitoring. Int. Arch. Occup. Environ. Health 62: 471–477 (1990). 2. E.J. Cone, R. Lange, and W.D. Darwin. In vivo adulteration: excess fluid ingestion causes false-negative marijuana and cocaine urine test results. J. Anal. Toxicol. 22: 460–473 (1998). 3. DHHS. Mandatory guidelines and proposed revisions to mandatory guidelines for federal workplace drug testing programs; Notices. Fed. Regist. 69: 19644–19673 (2004). 4. U. Mareck, H. Geyer, U. Flenker, T. Piper, M. Thevis, and W. Schanzer. Detection of dehydroepiandrosterone misuse by means of gas chromatography–combustion-isotope ratio mass spectrometry. Eur. J. Mass Spectrom. (Chichester, Eng) 13: 419– 426 (2007). 5. N. Baume, L. Avois, C. Schweizer, C. Cardis, J. Dvorak, M. Cauderay, P. Mangin, and M. Saugy. [13C]Nandrolone excretion in trained athletes: interindividual variability in metabolism. Clin. Chem. 50: 355–364 (2004). 6. World Anti-Doping Agency. Reporting and Evaluation Guidance for Testosterone, Epitestosterone, T/E Ratio and Other Endogeneous Steroids. WADA Technical Document - TD2004EAAS: 111, 2004. 7. M. Carrieri, A. Trevisan, and G.B. Bartolucci. Adjustment to concentration-dilution of spot urine samples: correlation between specific gravity and creatinine. Int. Arch. Occup. Environ. Health 74: 63–67 (2001). 8. T.E. Barber and G. Wallis. Correction of urinary mercury concentration by specific gravity, osmolality, and creatinine. J. Occup. Med. 28: 354–359 (1986). 9. L. Alessio, A. Berlin, A. Dell’Orto, F. Toffoletto, and I. Ghezzi. Reliability of urinary creatinine as a parameter used to adjust values of urinary biological indicators. Int. Arch. Occup. Environ. Health 55: 99–106 (1985). 10. L. Levine and J.P. Fahy. Evaluation of urinary lead determinations. J. Ind. Hyg. Toxicol. 27: 217–223 (1945). 11. E.J. Cone, R. Jufer, and W.D. Darwin. Forensic drug testing for opiates. VII. Urinary excretion profile of intranasal (snorted) heroin. J. Anal. Toxicol. 20: 379–391 (1996). 12. E.J. Cone, Y.H. Caplan, D.L. Black, T. Robert, and F. Moser. Urine drug testing of chronic pain patients: licit and illicit drug patterns. J. Anal. Toxicol. 32: 530–543 (2008). 13. B.A. Goldberger, B. Loewenthal, W.D. Darwin, and E.J. Cone. Intrasubject variation of creatinine and specific gravity measurements in consecutive urine specimens of heroin users. Clin. Chem. 41: 116–117 (1995). 14. M.A. Huestis and E.J. Cone. Differentiating new marijuana use from residual drug excretion in occasional marijuana users. J. Anal. Toxicol. 22: 445–454 (1998). 15. A. Trevisan. Concentration adjustment of spot samples in analysis of urinary xenobiotic metabolites. Am. J. Ind. Med. 17: 637–642 (1990). 16. J.E. Haddow, G.J. Knight, G.E. Palomaki, L.M. Neveux, and B.A. Chilmonczyk. Replacing creatinine measurements with specific gravity values to adjust urine cotinine concentrations. Clin. Chem. 40: 562–564 (1994). 17. R.C. Miller, E. Brindle, D.J. Holman, J. Shofer, N.A. Klein, M.R. Soules, and K.A. O’Connor. Comparison of specific gravity and creatinine for normalizing urinary reproductive hormone concentrations. Clin. Chem. 50: 924–932 (2004). 18. J.D. Cook, Y.H. Caplan, C.P. LoDico, and D.M. Bush. The characterization of human urine for specimen validity determination in workplace drug testing: a review. J. Anal. Toxicol. 24: 579–588 (2000).

Manuscript received April 18, 2008; revision received June 9, 2008.

7

Suggest Documents