Phosphatidylethanol in Blood as a Marker of Chronic Alcohol Use: A Systematic Review and Meta-Analysis

1Alcohol abuse and dependence are globally prevalent disorders, which span various socio-demographic groups and produce a broad range of secondary injuries and diseases.

2Objective clinical and biochemical tests for characterizing the drinking pattern, quantifying the amount of daily ingested ethanol, and efficiently detecting alcohol-induced disorders are critically needed in both treatment and research areas.

3Guidance on the investigation of suspected alcohol dependence or abuse includes symptoms, medical history, self-report forms, special questionnaires, clinical examination and biochemical investigations. Detailed efforts have been made to construct interview formats that correctly quantify alcohol intake, such as the “Alcohol Use Disorders Identification Test Consumption” (AUDIT-C), the CAGE questionnaire, or which include reports from collateral individuals. These approaches exhibit, however, limitations in the forensic setting, where individuals are motivated to deny or minimize the magnitude of their drinking behavior in order to mitigate the professional and legal ramifications of alcohol abuse.

4The limited diagnostic efficiency of self-reports and the difficulty in assessing alcohol-drinking behavior from an objective point of view have led in the last decades to an intensive search for reliable biomarkers of chronic excessive drinking; these markers can be broadly divided into direct and indirect categories. Indirect biomarkers detect the effects of alcohol on organ systems or body chemistry, and generally comprise markers of alcohol-related toxic effects, including mean corpuscular volume (MCV), aspartate aminotransferase (AST), alanine aminotransferase (ALT) and gamma-glutamyltransferase (GGT). More recent indirect markers examine ethanol-related biochemical changes in serum and comprise N-acetyl-beta-hexosaminidase (Beta-HEX), salsolinol, 5-hydroxytryptophol (5-HTOL), plasma sialic acid index of apolipoprotein J (SIJ) and the widely used carbohydrate deficient transferrin (CDT). Direct markers include blood ethanol itself, as well as alcohol derivatives, such as acetaldehyde, acetic acid, fatty acid ethyl esters (FAEE), ethylglucuronide (EtG), ethyl sulphate (EtS), and phosphatidylethanol (PEth).

5Among them, PEth in blood, and FAEE or EtG in hair, have attracted special attention as they are postulated to be highly specific and to roughly correlate with the ingested ethanol dose. Differently from FAEE and EtG, PEth in blood also seems to be promising for characterizing the drinking pattern (i.e., identifying binge drinking episodes) and differentiating light-moderate drinking from abstinence.

6For the above-mentioned reasons and given that all the identified review articles on the use of PEth in blood are based on descriptive data, a systematic review was conducted in order to summarize and better evaluate the diagnostic effectiveness of this marker in different clinical settings.

7As reported in Figure 1, the combined search with both MeSH terms and free-text protocols in the databases PubMed, Web of Science, Scopus and Ovid/Embase retrieved 916 records, 472 of which were excluded, as they were duplicates. Of the 444 records screened by title and abstract, 386 were manually excluded, because they did not fulfill the inclusion criteria. In particular, in those manuscripts, PEth was used for characterizing the function and properties of the enzyme phospholipase D, not as a biomarker of chronic alcohol consumption.

8Fifty-eight (58) potentially relevant papers were analyzed in full-text (Figure 1) and used to summarize the current evidence on:

9– The formation, distribution and degradation of PEth in human blood (Section 2.1);

10– The presence and distribution of different PEth molecular species in human blood (Section 2.2);

11– The most diffused and efficient analytical methods devoted to the identification and quantisation of PEth in human blood (Section 2.3);

12– The diagnostic efficiency of PEth as a clinical marker of chronic excessive drinking (Section 2.4);

13– The potential utility of PEth as a marker of heavy episodic drinking or binge drinking (Section 2.5).

14Twelve (2.7%) studies were included in the meta-analysis (Tables 1 and 2; Figure 2) since they presented integrable data on PEth concentration in human blood, and thus on the diagnostic efficiency of the marker in detecting harmful drinking behaviors. The mean and 95% confidence interval (CI) of total PEth concentration in social drinkers (mean 0.288 μM; CI 0.208–0.367 μM) and heavy drinkers (mean 3.897 μM; CI 2.404–5.391 μM) were reported in Figure 2, and discussed in Sections 2.4 and 2.5.

15Phosphatidylethanol (PEth) is an abnormal cellular membrane phospholipid and was discovered for the first time in mammalians in 1983, being detected in the brain, kidney, liver, skeletal muscle, and heart of rats chronically exposed to ethanol.

16The formation of PEth is catalyzed by phospholipase D (PLD), an ubiquitary enzyme normally devoted to the hydrolysis of phosphatidylcholine (PC) to phosphatidic acid (PA). PLD has a high affinity for short chain alcohols (100–1000-fold higher than for water). In the presence of ethanol, it promotes a transphosphatidylation reaction, with the production of PEth. The expression of PLD varies dramatically among different animal species and also among the tissues of a single organism. In humans, two different isoforms of the enzyme (PLD1 and PLD2), sharing 50% of the DNA sequence, have been genetically and functionally characterized. PLD1 has a perinuclear distribution and displays a very low basal activity, requiring a protein kinase C activation; PLD2 is localized in the cellular membrane and is constitutively active. Both PLD1 and PLD2 catalyze the formation of PEth in human red blood cells.

17It has been observed that the in vitro incubation of whole human blood with ethanol for 24 h (EtOH concentration varying between 50 and 100 mM) induces the formation of PEth, and that the EtOH concentration and the incubation time are directly proportional to the quantity of PEth produced. An activator of the protein kinase C added to the mixture triplicates the generated quantity of PEth. No correlation between hematological indexes (red blood cell count, mean corpuscular volume, hematocrit) and the rate of PEth formation has been found.

18Several studies on blood collected from mice, rat, ferret and pig have excluded the presence and the in vitro production of PEth in animal red blood cells (RBCs); thus, human RBCs seem to be peculiar in forming PEth in vitro in the presence of ethanol. This characteristic represents an important drawback of PEth as a marker of chronic alcohol abuse: samples collected when the blood ethanol concentration (BAC) is higher than 0.1 g/L can generate false positive results due to the neo-formation of PEth in vitro (in the post-sampling period), which can occur at room temperature (16 °C–20 °C), but also at −20 °C, being slower at +4 °C. Only at −80°C can the formation process be considered drastically inhibited.

19Another important and peculiar characteristic of human RBCs is the incapacity of efficiently degrading PEth, probably due to the absence of phosphatidylcholine phospholipase C (PLC) activity. PEth elimination, with a half-life (t/2) varying between 0.5 and 2 h, has been demonstrated in several human cellular systems, such as pancreatic islets, hepatocytes (HepG2 and C6 cells), and neutrophils, but not in RBCs. The molecular mechanism of PEth elimination has not yet been fully elucidated; although phospholipase A2 and PLC in vitro release arachidonic and palmitic acid from PEth, there is still a lack of evidence that phospholipases are involved in the in vivo PEth degradation. What is clear is that in human RBCs there is a disproportion between PEth formation and degradation rates, causing PEth to accumulate in the cellular membranes, an accumulation that suggests a potential use of PEth for detecting chronic exposure to ethanol.

20In clinical studies conducted on chronic heavy drinkers, PEth was found to be detectable in blood up to 28 days after sobriety. In 15 alcoholics following a detoxification program, the mean half-life of blood PEth was 4.0 ± 0.7 days with a range of 3.0–5.3 days. The kinetics of elimination was well-approximated by a one-compartment model. More recently, this degradation kinetics has been confirmed on 57 alcohol-dependent subjects following a detoxification program. PEth decreased over time with a half-life of about 3–5 days and was detectable in 64.3% of the cases after 28 days of sobriety. Additionally, it has been demonstrated that sex, gender, age and body mass index do not influence the normalization rate of PEth.

21In a recent experiment during which, after three weeks of abstinence, 11 social drinkers were exposed to an amount of ethanol of 1 g/Kg for five consecutive days (daily alcohol intake ranging between 67 and 109 g/die), and then remained abstinent for 16 days, undergoing regular and scheduled blood sampling, the mean half-life of PEth ranged from 4.5 to 10.1 days in the first week and from 5.0 to 12.0 days in the second week.

22Since the very early studies conducted on rats chronically exposed to ethanol it has been observed that PEth is not a single molecule, but a group of glycerophsopholipid homologues with a common phosphoethanol head group onto which two long carboxylic acid side chains, typically containing from 14 to 22 carbon atoms with different grades of insaturation (0–6 double bonds), are attached.

23These homologues are commonly named in the form “PEth A:B/C:D” where A and C indicate the number of carbons in the carboxylic side chains, whereas B and D indicate the number of double bonds in each side chain.

24Fast atom bombardment-mass spectrometry experiments on PEth, formed after bradykinin or phorbol ester PLD stimulation in pheocromocytoma cells (PC12), revealed that the molecular species of the generated PEth were almost identical to those of PC, consistent with this lipid being the substrate of PLD. Additionally, using neuroblastoma cells (NG 108-15) it has been demonstrated that the addition of polyunsaturated fatty acids to the medium induced similar changes in the fatty acid composition of PC and PEth.

25In one of the studies by Alling et al. on the distribution of PEth in the organs of rats chronically treated with ethanol, significant inter-organ differences in the fatty acid composition of PEth have been highlighted: brain PEth contained high proportions of palmitic (16:0) (28%) and oleic (18:1) (34%) acids with a 10% of polyunsaturated fatty acids (mainly arachidonic acid, 20:4), whereas liver PEth contained more stearic acids (18:0) and less palmitic acids (16:0).

26Only in recent times was the characterization of PEth molecular species in human blood performed, through highly sensitive liquid chromatography multiple mass spectrometry methods (LC-MS/MS). In blood collected from heavy drinkers PEth 16:0/18:1 and 16:0/18:2 have been demonstrated to be the predominant molecular species accounting on average for 37%–46% and 26%–28%, respectively, of total PEth in blood. Owing to inter-individual variations, PEth 16:0/18:2 was sometimes the major form, whereas PEth 16:0/20:4 constituted about 8%–13% of total PEth, and PEth 18:1/18:1 and 18:0/18:2 taken together constituted about 11%–12% of total PEth. The species 16:0/16:0, 16:0/20:3, 18:0/18:1, 16:0/20:3 and 18:1/18:2 had a lower blood concentration, each one accounting for about 1%–5% of total PEth. Diet was suggested to be the main factor determining the fatty acid pattern in PC and PEth, although the influence of genetic determinants, drinking pattern and metabolic disorders needs to be further clarified.

27The numerous combinations of chain length and double bonds enable the formation of a very large theoretical number of different PEth species and so far 48 homologues have been identified in post-mortem human blood. Although inter-individual differences in the relative abundances of PEth homologues do exist, the preliminary data available suggest that 5 molecular species (16:0/18:1, 16:0/18:2, 16:0/20:4, 18:1/18:1, 18:1/18:2) could constitute more than 80% of total PEth.

28For PEth determination, venous blood should be collected in tubes containing ethylenediamine tetra-acetic acid (EDTA) and the sample should not be centrifuged. Blood samples for PEth analysis have proved to be stable for 24 h at room temperature and for 3 weeks at +4 °C. For longer periods of storage, whole blood should be frozen in a plastic tube and kept at −80 °C, to avoid any in vitro formation.

29Several analytical strategies have been utilized so far for quantifying total PEth concentration in blood, mainly based on chromatographic or electrophoretic separation methods, as reported below.

30– Thin layer chromatography (TLC).

31– High performance liquid chromatography (HPLC) coupled to an evaporative light-scattering detector (ELSD).

32– Gas-chromatography coupled to mass spectrometry.

33– Non-aqueous capillary electrophoresis (NACE).

34– Immunoassay with PEth-specific monoclonal antibodies.

35Thin layer chromatography (TLC) is a manual semi-quantitative method with limited sensitivity and throughput and was the very first method used for identifying PEth in animal tissues, quickly judged unsuitable for further experimental studies. The first and only gas-chromatographic method reported in the literature was developed for the detection of two derivatization products of PEth, ethyl bis (trimethyl-silyl)-phosphate and tris (trimethyl-silyl)-phosphate; probably due to a lack of specificity of the monitored derivatization products, this method has never been applied to the analysis of clinical samples.

36Similarly, to the best of our knowledge, the immunochemical method based on the anti-PEth antibodies 2B1 e 2E9 generated with the traditional hybridoma technique, capable of detecting in vitro formed PEth, has not yet found a clinical application.

37The two electrophoretic methods based on non-aqueous capillary electrophoresis, either coupled to an UV or to a mass spectrometric detector have also found very limited clinical application.

38On the contrary, the HPLC-ELSD method is actually the most utilized in clinical toxicology. Briefly, whole blood together with the internal standard (phosphatidylbutanol 18:1/18:1) are extracted with 2-propanol and hexane, followed by the quantification of PEth in the extract on an HPLC system equipped with an evaporative light scattering detector (ELSD) and with PEth 18:1/18:1 as calibrator.

39More recently several liquid chromatography mass spectrometry (LC-MS) or multiple mass spectrometry methods (LC-MS/MS) have been developed for the identification of PEth homologues in blood, the majority of which employ electrospray ionization (ESI) for MS coupling, although time-of-flight (TOF) mass spectrometry has also been proposed. Also a LC-MS/MS method for the analysis of PEth 16:0/18:1 and 18:1/18:1 on dried blood spots, which exhibited a good correlation with parallel determinations on fresh blood, has been developed and validated.

40All these novel mass spectrometric methods exhibit a considerably higher analytical sensitivity (more than two orders of magnitude) and shorter turnaround time with respect to HPLC-ELSD, and are capable of identifying PEth species also in blood collected from social drinkers.

41A considerable limitation of the above-mentioned methods, however, is the absence of commercially available reference substances for PEth analogs (at the moment only PEth 16:0/16:0, 16:0/18:1 and 18:1/18:1 are available), complicating the validation process, and hindering the diffusion of these methods in clinical and forensic toxicology laboratories. Zheng and coworkers have solved the problem by preparing in house deuterium-labeled PEth analogs for PEth 16:0/18:1 and PEth 16:0/18:2. This commendable effort is however too complex and time-consuming to leave the research arena and enter routine laboratory practice.

42Four out of twelve articles included in the present meta-analysis are clinical trials involving inpatients with a diagnosis of alcohol-dependence based on the Diagnostic and Statistical Manual IV-R (DSM IV-R) or the International Classification of Diseases (ICD-10) criteria, who underwent a detoxification program (see Table 1). These studies are generally aimed at evaluating the diagnostic efficiency of PEth in comparison to clinical interviews, questionnaires and/or other traditional markers of harmful drinking, and at correlating the concentration of PEth to the amount of ethanol ingested in the previous 2–4 weeks. Only one of those studies used an intergroup control, the rest being uncontrolled trials.

43The concentrations of total PEth in alcohol dependent subjects admitted for detoxification vary significantly in the considered clinical studies (range: 0.0–7.7 μM; see Table 2). With respect to the mean values and 95% CI of PEth concentrations reported in the meta-analyzed papers (see Table 2 and Figure 2), the heavy drinkers group (Daily Alcohol Intake—DAI > 60 g) is well separated from the social drinkers, displaying a mean concentration one order of magnitude higher than the remaining groups (heavy drinkers = 3.897 μM; social drinkers = 0.288 μM).

44At the present time, the international scientific community has not yet established a cut-off value for PEth concentration in blood to be used for differentiating an acceptable social ethanol intake (<40 g for males and <20 g for females, according to the World Health Organization parameters), from an at-risk-alcohol-use (40–60 g/die) and chronic excessive drinking behavior (>60 g/die).

45Nine of the above-mentioned clinical trials used HPLC-ELSD for total PEth quantification in blood utilizing as an interpretative cut-off the lower limit of quantification (LOQ) of the analytical method used for PEth determination, that is: 0.22 μM, 0.30 μM or 0.8 μM. With the above-mentioned cut-offs, the diagnostic sensitivities of PEth were very high, varying from 98% to 100%.

46In Sweden, 0.7 μM is currently used as the routine clinical threshold, although that value has not yet been approved at a supranational level.

47Regarding clinical specificity in differentiating alcohol-dependent subjects from social drinkers and/or abstainers, the available controlled trials have all obtained a 100% value. This absolute specificity of the biomarker can only be partly explained by the intrinsic characteristics of PEth, which is formed in blood only in the presence of ethanol. Potential bias of selection have to be considered, due to the fact that the case-control populations considered exhibited significant differences in the drinking pattern and in the amount of daily-ingested ethanol (see Table 1).

48Differing from the traditional indirect biomarkers used for diagnosing a chronic excessive drinking behavior (MCV, AST, ALT, GGT, and CDT), blood PEth concentration seems not to be influenced by age, gender, other ingested substances or non alcohol-associated diseases, such as hypertension, kidney and/or liver diseases. For these reasons, PEth is considered to perform better than MCV and GGT, both in terms of sensitivity and specificity, for detecting chronic excessive drinking behaviors. A few studies have also highlighted a slightly higher sensitivity of PEth, compared to the most selective indirect marker CDT.

49Preliminary evidence does exist concerning a rough correlation between the amount of ethanol consumed in the previous two weeks and the concentration of PEth in blood, although observations of individual PEth formation rates and recent studies on moderate drinkers indicate that it might not always be possible to link PEth concentration in blood to a precise drinking level.

50An especially interesting question, which still needs to be answered, regards the quantity of ethanol that must be consumed for a certain time-period to give a positive blood PEth assay. Using the HPLC-ELSD method of analysis, it has been observed that a single ethanol dose of 30–47 g did not produce any measurable amounts of PEth in blood. The threshold of total ethanol intake leading to a positive PEth assay was estimated at around 1000 g in three weeks, with a daily consumption of at least 50 g/die. A recent drinking experiment, which employed a more sensitive LC-MS/MS method for the quantification of PEth 16:0/18:1, conducted on 11 healthy volunteers who drank 50–109 g of ethanol/die, showed that the formation of PEth began immediately after the first assumption of alcohol (0.5–8 h) reaching a concentration of about 0.05–0.10 μM and then stagnated or decreased when the blood alcohol concentration started to decline. These results are certainly encouraging because they pave the way for novel potential applications of PEth in the diagnosis of excessive drinking episodes and/or “binge drinking” behaviors; at the same time, they do underline the compelling need to study large populations of social drinkers and teetotalers in order to determine whether an efficient cut-off can be established for differentiating teetotalers from social/moderate drinkers and “binge drinkers”.

51At the present time, only five studies have examined the characteristics of total PEth as a potential marker of “binge drinking”; the small number of the investigated subjects, the variability of the clinical methods used to reconstruct the daily alcohol intake (see Table 1), and the diversity of the analytical methods employed for quantifying total PEth in blood (see Table 2) has not yet allowed any definitive conclusion to be drawn. Comasco et al. have found a sensitivity of only 9% for detecting a moderate alcohol consumption (>2 drinks/week) in 200 adolescent students from Vestmanland, with the limited clinical efficiency of PEth probably explained by the low sensitivity of the analytical method (HPLC-ELSD) and the subject stratification employed.

52Stewart and colleagues, examining 80 healthy women in reproductive age, have found that a PEth 16:0/18:1 concentration above 0.18 μM was highly specific for identifying women drinking more than 28 g per day, although a relevant degree of inter-individual variability was evident; this variability could only partly be explained by the different timing of assumption (with respect to blood sampling) and the different drinking patterns (regular moderate intake vs. episodic heavy drinking). Although being expensive, we do believe that only controlled drinking experiments over a period of 2–3 weeks could highlight if and how PEth could be utilized as a marker of “binge drinking” or as a marker of absolute abstinence.

53In March 2012, one of the authors (GV) performed the systematic search of the literature searching Medline/PubMed, SCOPUS, Web of Science, and Ovid/Embase databases. The Medline search employed a complex search strategy including both “MeSH” and “free-text” protocols. More specifically, the following terms retrieved from the MeSH browser provided by Medline were utilized: (“Ethanol” [MeSH] OR “Alcohol Drinking” [MeSH]) AND “Phosphatidylethanol” [MeSH]. A multiple “free-text” search restricted to the fields “title/abstract” was performed combining by “AND” the entry terms “Alcohol” and “Phosphatidylethanol”. No temporal limits were utilized. Because of the interface limitation only the “free-text” protocol “((Phosphatidylethanol OR PEth) AND (Ethanol OR Alcohol))” was used for searches in Ovid/Embase, Web of Science (search field: “topic”), and SCOPUS (search fields: title/abstract, keywords).

54Paper selection was conducted independently by three reviewers (GV, RBB, AN), based on titles and abstracts of papers retrieved by the systematic search. The following inclusion and exclusion criteria were adopted for review and meta-analytic purposes. Any discrepancy in the paper selection and data extraction was settled by consensus discussion.

55To meet the inclusion criteria for the present review, studies had to fulfill at least one of the following requirements:

56Provide data on PEth concentration in human blood collected from alcohol dependent subjects, heavy drinkers, moderate drinkers, social drinkers or teetotalers.

57Provide data on PEth concentration in fluids or tissues of animals chronically exposed to ethanol.

58Describe an analytical method for the quantification of total PEth or PEth molecular species in human blood.

59Articles not fulfilling at least one of the previous requirements or characterizing the activity of phospholipase D (PLD) were excluded.

60In the case of doubtful classification based on solely title and abstract, the full text was retrieved. Whenever this was ineffective, the question was settled by consensus discussion.

61To meet the inclusion criteria for the present meta-analysis, studies had to fulfill all the following requirements:

62(D) Fulfilling criterion A for Review purposes.

63(E) Reporting integrable data on blood concentrations of PEth in humans.

64(F) Reporting integrable data on the daily alcohol intake (DAI) of the subjects recruited in the study.

65Articles not fulfilling all of the above-mentioned requirements, being commenting letters or reviews, reporting data not comparable/suitable for direct meta-analysis processing (i.e., inappropriate statistical formats) or not amenable by extraction or calculation/conversion from the published results or figures, were excluded.

66Data extraction was conducted independently by four authors (GV, RBB, AN, PF) and the data derived from the studies were collected in an electronic database, while two different authors (GC, SDF) verified the accuracy of the data extraction process, in order to minimize subjective judgment. The following items were collected from each study: authors, publication year, features of the study (main aims, inclusion and exclusion criteria, duration of the follow-up), characteristics of the investigated population (numbers of subjects, age, race, comorbidities, use of medications or assumption of illicit drugs, clinical setting, type of controls used), type and amount of alcohol consumption (estimation of alcohol use, daily mean ingested ethanol, type of stratification if any, timing of sample collection, biological assessments before collection), analytical methods used for PEth analysis, type of measured PEth and concentrations, type of collected sample, clinical efficiency of PEth (sensitivity, specificity, positive predictive value, negative predictive value), and other markers used for detecting chronic excessive drinking. Any discrepancy in data extraction was settled by consensus discussion.

67In order to meta-analyze the distribution of PEth concentration with regard to the DAI, data were grouped into two categories, as follows, considering the mean or median of DAI for the classification of the drinking behavior.

68Category 1. Social drinkers (DAI ≤ 60 g).

69Category 2. Heavy drinkers (DAI > 60 g).

70Specifically, these categories were overlapping for some extreme values (ranging from 40 to 60 g), which represented the mathematical limits of the Standard Deviation or the Interquartile Range. This reflected the lacking of a shared threshold to discriminate DAI among groups. Nevertheless, the descriptive statistics of DAI also reported the most important indexes of central tendency (mean and median), which were markedly different between groups (Table 2), hence not impairing our analysis.

71One of the authors (RBB) performed the meta-analysis according to a previously published procedure. Statistical analyses of continuous variables were performed using the weighted mean as the summary statistic reported with 95% confidence intervals. To allow inference to an external population, a random-effects model was used for purposes of meta-analysis in a conservative setting.

72The present systematic review demonstrates that total phosphatidylethanol, an abnormal phospholipid formed in the erythrocyte membrane exclusively in the presence of ethanol, exhibits high diagnostic sensitivity and specificity for detecting active chronic excessive drinking behaviors, with a regular daily alcohol intake (DAI) of more than 60 g.

73The mean values and confidence intervals of total PEth concentrations in blood of heavy (DAI > 60 g/die) and social drinkers (DAI ≤ 60 g/die) showed a significant statistical difference. These findings demonstrate a good clinical efficiency of PEth for detecting heavy drinking.

74The recent introduction of sensitive analytical methods based on liquid chromatography coupled to mass spectrometry detection, capable of effectively measuring single molecular species of PEth in blood in the nanomolar range, has opened promising new application fields for PEth, such as the identification of minute alcohol consumption (i.e., monitoring of abstinence), and the identification of heavy episodic drinking behaviors (i.e., “binge drinking”). However, large randomized trials are needed in order to ascertain if PEth is really effective in those diagnostic challenges.

75Moreover, it is essential to set up international standards that properly define the characteristics of the alcohol biomarker PEth and harmonize the methodology used in both clinical and laboratory investigations. The following five issues certainly need to be addressed.

76Define the molecular species of PEth that should be determined in blood when a mass spectrometric analytical method is used; it needs to be clarified if PEth 16:0/18:1 and 16:0/18:2 (the two most abundant homologues) are sufficient to resemble the total PEth concentration or whether it would be more appropriate to investigate and quantify multiple PEth molecular species; the data derived from our analysis suggest that at least 5 molecular species (PEth 16:0/18:1, 16:0/18:2, 16:0/20:4, 18:1/18:1, 18:1/18:2) could be used.

77Develop and market commercial reference substances for the PEth molecular species identified by point 1.

78Arrive at a consensus on the most appropriate cut-off for differentiating social alcohol use from heavy drinking.

79Characterize the kinetics of formation and degradation of the identified PEth species in order to determine whether they might be used for monitoring abstinence.

80Verify the correlation of blood PEth concentration with the amount of ethanol ingested in the previous two weeks.