Testing for Heavy Use of Alcohol Scott H. Stewart, MD Center for Drug and Alcohol Programs, Charleston Alcohol Research Center Department of Psychiatry and Behavioral Sciences
Epidemiology of Alcohol Use and Alcohol-Associated Problems Estimates from National Epidemiologic Survey on Alcohol and Related Conditions (NESARC), 2001-2002 and World Health Organization Studies on Health Risks
Current Drinking (any in past year) Women %Current drinkers %Current drinkers that drank heavily Men Non-Hispanic white 65.1 (0.79) 13.9 (0.47) Non-Hispanic black 45.9 (1.01) 12.7 (0.96) Hispanic 49.5 (1.51) 8.8 (0.92) Non-Hispanic white 74.3 (0.73) 18.5 (0.55) Non-Hispanic black 62.6 (1.25) 19.9 (1.10) Hispanic 70.0 (1.20) 13.8 (1.04)
Alcohol Use Disorders Among Current Drinkers Group Alcohol Dependence (%) Women Non-Hispanic white 3.6 (0.3) Non-Hispanic black 5.2 (0.7) Hispanic 3.9 (0.6) Men Non-Hispanic white 7.3 (0.4) Non-Hispanic black 8.1 (0.8) Hispanic 8.4 (1.0)
Alcohol and All-Cause Mortality Abstainers 1-2 drinks/day Heavy use Nadir due to a decrease in ischemic heart disease death Subsequent increase mainly due to accidents (~50%), cancers (~20%), and liver disease (~14%), as well as other alcohol-associated conditions
Alcohol and Disease Burden Disability adjusted life years Metric that incorporates years of life impaired by disability as well as years of life lost Estimated from population-level average consumption and drinking patterns and alcohol-attributable fractions for injury and disease Alcohol in Americas-A (US and Canada) Accounts for 11.9% of DALY s among males Accounts for 3.2% of DALY s among females Accounts for 2% of deaths among males Overall protection among females due to projected decrease in ischemic heart Rehm et al, Comparative Quantification of Health Risks, World Health Organization 2003
Examples of Causal Associations ~10% of hypertension caused by alcohol BP up with drinking in excess of 2 drinks per day End organ disease (e.g., liver, heart, brain, etc.) Malignancies (oropharynx, UGI tract, liver, others) Alcohol-related disorders in current DSM Delirium, dementia, amnestic disorder, psychotic disorder, mood disorder, anxiety disorder, sleep disorder Almost half of hospitalized trauma victims (LM Gentilello et al, JAMA, 1995)
If We Need to Know About Potentially Unhealthy Drinking, Is Self-Report Sufficient?
Reasons for Under-reporting Alcohol Use Drinking in moderation largely accepted in Western society, but chronic heavy drinking and impaired control is stigmatized Some reasons for under-reporting include: Social context (e.g., contingencies associated with particular responses, interpersonal rapport) Respondent characteristics (e.g., attitudes, values, beliefs) Task attributes (e.g., clarity of instruction, instrument design) Del Boca and Darkes, Addiction 2003, 98 (suppl 2): 1-12
Proportion of Baseline Median Value Mismatch Between Self-Report and Phosphatidylethanol Over Time 0 3 6 10 16 Alcohol Dependence Treatment Week
Alcohol Consumption Biomarkers, Focus on Carbohydrate-Deficient Transferrin, Phosphatidylethanol, and Ethyl Glucuronide
Why Focus on These Three? PEth and EtG require ethanol for synthesis (i.e., specific for some ethanol exposure) PEth in blood has a 1 to 4 week horizon, but does not require heavy drinking for detection EtG in hair detects moderate to heavy drinking in prior months CDT Greater specificity for chronic, heavy drinking in prior weeks, particularly relative to other toxic effects markers Particularly well validated
Distribution of Ethanol Ingested ethanol is rapidly absorbed in the stomach (~20%) and duodenum (~80%) Roughly 10% of ethanol (more with food) is quickly metabolized in the liver and does not reach systemic circulation ( first pass ) Readily crosses cell membranes Distributed throughout the body with tissue concentration dependent on water content Quickly (seconds to minutes) equilibrates with plasma ethanol Arthur Cedarbaum, RSA Lecture Series (www.rsalectures.com/slides.html)
Elimination of Ethanol Average elimination capacity is around 10 grams ethanol/hour About 3-4 fold variation in capacity Major mechanism is oxidation in liver ADH and ALDH (mostly concentration independent) Cytochrome P450-2E1 (inducible, concentration dependent) Liver bears brunt of ethanol, with physiologic effects (CDT) Some (<10%) directly eliminated through lungs, sweat, urine Small proportion eliminated by several minor pathways (glucuronidation (EtG), transphosphatidylation (PEth)) Arthur Cedarbaum, RSA Lecture Series (www.rsalectures.com/slides.html)
Carbohydrate-Deficient Transferrin
Carbohydrate-Deficient Transferrin First described in 1978 Abnormal protein band with IEF in serum (and CSF) of 16 alcoholic patients found to be transferrin variant Normalized after 10+ days of abstinence Relative deficit of sialic acid and then galactose reported, eventually leading to the term carbohydrate-deficient transferrin Research on the use of CDT as an alcohol biomarker quickly commenced and continues Specific molecular form identified mid 90 s to early 00 s
Carbohydrate-Deficient Transferrin Increase in CDT induced by drinking ~50 to 60 grams/day for at least one or two weeks (i.e., about 4 or more drinks) About 60% of such drinkers have a positive assay at cutoffs providing ~95% specificity Half-life is ~1.5 to 2 weeks, and can therefore be used to assess response to treatment or detect relapse after a period of abstinence
What is it? Heavy Drinking Can Alter Protein Glycosylation N-glycosylation Attachment of oligosaccharide precursor to asparagine residues (Asn-X-Ser/Thr) Process occurs in the rough endoplasmic reticulum and Golgi apparatus Involves secreted or transmembrane proteins Chronic heavy drinking is known to decrease N-glycosylation of some hepatically synthesized serum proteins
Altered Transferrin Glycosylation Alcoholism can induce production of transferrin with reduced N- glycosylation Alcohol-induced loss of one or both N- glycosylation sites (which occurs normally to a lesser extent) can be quantified Image from Essentials of Glycobiology (2 nd ed). Cold Spring Harbor Laboratory Press, 2009
Alcohol-Induced Carbohydrate Deficient Transferrin is Disialotransferrin Absence of heavy drinking Heavy drinking
Transferrin Chromatogram
The importance of assay and analyte Isoelectric focusing-western blot Charge-based separation of iron-saturated transferrin isoforms and immunofixation Early tests were highly valid but cumbersome Susceptible to overlap with other isoforms which may not separate well from disialotransferrin Small anion-exchange column to separate transferrin isoforms followed by transferrin immunoassay (e.g., CDTect, %CDT TIA) Commercialized but less selective for disialotransferrin Some trisialotransferrin also measured with potential false positive results Mass Spectrometry Ratio of aglycosylated and disialotransferrin to tetrasialotransferrin Genetic variants likely identified Effects of liver disease may not be noted
Targeted CDT Assays N-Latex immunoassay Antibody derived from recombinant transferrin (i.e., no N- glycans) Detects asialo, monosialo, disialo transferrin relative to total transferrin High performance liquid chromatography Selective for disialotransferrin Genetic variants that may invalidate test are identified Effects of liver disease that may invalidate test are identified Capillary electrophoresis Similar to HPLC
CDT- False Positive/Negative? With older commercial kits, additional confirmation is important Liver disease, hepatocellular cancer, certain genetic variants of transferrin, congenital disorders of glycosylation (rare) With HPLC method, much less of an issue
Use of CDT To rule in regular heavy drinking 95+% specificity Ideally, final determination of positive should include testing that resolves and quantifies disialotransferrin To monitor for response to treatment or for relapse Half life 1.5 to 2 weeks Rule out chronic heavy drinking? Remember that a substantial minority of chronic heavy drinkers will not exceed the cutoff (~30 to 40%)
Phosphatidylethanol
Minor Ethanol Elimination Route Phosphatidylcholine + Ethanol PLD Phosphatidylethanol Embedded in plasma membrane of erythrocytes (and elsewhere).
Phosphatidylethanol Discovered serendipitously by two groups in 1983 (Madison, WI and Amsterdam) Clinical research literature focusing on PEth as an alcohol biomarker began appearing in the mid to late 1990 s Preliminary studies exclusively European Two main assays (seem to perform similarly for heavy drinking): HPLC-ELSD (quantifies family of PEth s) LC-MS/MS (quantifies single species/isomer)
Phosphatidylethanol PEth accumulates in erythrocyte membrane (and probably other cell membranes as well) Half life is around 4 to 5 days Can be detectable for up to a month or more following the cessation of heavy drinking Will become undetectable more quickly in moderate drinkers Not expected to be detectable with occasional light drinking Levels modestly correlate with total alcohol intake in the past 1 to 2 weeks (r 0.6)
Early PEth Studies Citation Detection Method Cutoff/L OQ Cases (male) Definition of Case Controls (male) Definition of Control Sensitivity Specificity Correlation with EtOH Consumption Aradottir et al 2006 HPLC w/ ELSD >0.22 µmol/l 144 (123) Inpatient and outpatient EtOH dependent patients N/A N/A 99 N/A 0.568 (p<0.001) Hansson et al 1997 TLC >0.1 µmol/l 15 (15) 13 alcoholics admitted for detox, drinking 60-300g EtOH daily for 1wk prior to admission and 2 hospital admissions drinking 150-300g EtOH daily for 1wk prior to admission who were followed over time 6 (6) Non-heavy drinking lab personnel who had abstained for at least 4 days 100 100 Not reported Hartmann et al 2007 HPLC w/ ELSD >0.36 µmol/l 56 (not given) Alcohol dependent detox patients w/ median EtOH consumption of 1400g over past 7 days 35 (not given) Sober forensic psychiatric inpatients on a closed ward 95 100 0.802 (p<0.05) Wurst et al 2004 HPLC w/ ELSD 0.8 µmol/l 18 (14) Detoxification patients with an ICD-10 diagnosis of alcohol dependence N/A N/A 100 N/A Not reported Explanation of Abbreviations: HPLC w/ ELSD - High Pressure Liquid Chromatography with Evaporative Light-Scattering Detector LOQ - Limit of quantitation TLC - Thin Layer Chromatography
PEth in Non-Abstinent Reproductive Age Women PEth quantifiable in 57/80 subjects. This included: 8/15 drinking < 1 drink/day 23/37 drinking 1 to <2 drinks/day 26/28 drinking 2+ drinks/day PEth cutoffs: 45 ng/ml 95% specific and 61% sensitive for 1+ drinks/day 127 ng/ml 95 % specific and 40% sensitive for 2+ drinks/day
PEth (ng/ml) PEth and Past 14-Day Drinking 600.00 500.00 109 173 101 abstained 46 up to 2 drinks 44 > 2 drinks 400.00 300.00 181 200.00 100.00 156 168 165 0.00 Abstained Up to 2 drinks/day >2 drinks/day Average 2 Week Drinking by Self-Report
PEth (ng/ml) Change in PEth in Heavy Drinkers 450 400 350 300 250 200 150 100 50 0 Base 2 week 4 week
PEth- False Positives/Negatives? Unintentional exposures will not be great enough to result in detection Less than 2 drinks/day frequently not detected Transfusion of PEth-positive packed red blood cells could in theory cause a false positive (e.g., a trauma patient after initial resuscitation) If there is a small amount of ethanol in blood sample, RBC s may continue to synthesize PEth From 1 to 5% of heavily drinking individuals appear to metabolize little ethanol by this route Test + - Drinking Yes No
Use of PEth? Biomarker is not fully validated yet, but is specific for some alcohol use, and unintentional exposure will not result in detectable PEth (possible exceptions are invitro synthesis and blood transfusion) Suggested interpretation based on current knowledge: Detectable PEth by LC-MS or HPLC-ELSD indicates drinking roughly within the past 4 to 6 weeks Higher levels indicate heavy drinking within the past 2 to 3 weeks optimal cutoff concentrations not established Low levels indicate lighter recent drinking probably approaching one drink/day, or heavy drinking that ceased in the past 1 to 1.5 months
Ethyl Glucuronide
Minor Elimination Routes- Ethyl Glucuronide UDP-glucuronyltransferases in liver generates ethyl glucuronide COOH Ethanol + UDP-Glucuronate OH OH O OH CH2CH3
Ethyl Glucuronide Ethanol glucuronidation first detected in 1901 First detected in human urine in 1967 Clinical or judicial use has been reported since the mid 1990 s Much of the research in the clinical literature published from late 1990 s to present Main assay methods include MS and now immunoassays
Elimination of Ethyl Glucuronide Cleared from serum within hours of drinking cessation Mainly accumulates in urine and eliminated over 1 to 3 or 4 days Some accumulates in hair and nails, where it persists for months
Hair EtG (pg/mg) Hair EtG and Past 3-Month Drinking 300.00 250.00 173 85 abstained 58 up to 2 drinks 44 > 2 drinks 200.00 165 150.00 100.00 50.00 109 156 134 75 94 90 4 0.00 Abstained Up to 2 drinks/day >2 drinks/day Average drinking past 90 days by self-report
Use of Hair EtG? Not fully validated, but detectable hair EtG specific for some alcohol use probably averaging at least 1 drink/day in past months Suggested interpretation based on current knowledge: Detectable level likely indicates frequent drinking in past months Higher levels strongly suggest greater alcohol use, but valid cutoffs for heavy drinking not established
Final Thoughts
Limitations Blood PEth and hair/nail EtG specific for ethanol, but Not yet extensively validated May not be highly sensitive and specific cutoffs for heavy use Serum %dcdt False positives can occur (less of an issue with HPLC, capillary electrophoresis, and possibly N-Latex immunoassay and MS) Reasons for most false negatives are currently unknown Can t currently identify hazardous alcohol use in everyone (particularly infrequent binge drinking) But these markers will likely identify drinking that would be sufficient to contribute to chronic health conditions
Research Areas Pertinent to Risk Stratification for Alcohol Problems Discovery of biomarkers for early alcohol-related disease (e.g., dependence, liver or heart damage, etc.) Continued validation of current markers (e.g., determine optimal cutoffs, identify causes of false positives or negatives, identify factors that modify relationship between marker and alcohol consumption, etc.)