Assessment of alcohol consumption in liver transplant candidates and recipients: The best combination of the tools available
No grants supported this study.
The authors of this article have no conflicts of interest to disclose as described by Liver Transplantation.
The detection of alcohol consumption in liver transplant candidates (LTCs) and liver transplant recipients (LTRs) is required to enable a proper assessment of transplant eligibility and early management of alcohol relapse, respectively. In this clinical setting, urinary ethyl glucuronide (uEtG), the Alcohol Use Disorders Identification Test for Alcohol Consumption (AUDIT-c), serum ethanol, urinary ethanol, carbohydrate-deficient transferrin (CDT), and other indirect markers of alcohol consumption were evaluated and compared prospectively in 121 LTCs and LTRs. Alcohol consumption was diagnosed when AUDIT-c results were positive or it was confirmed by a patient's history in response to abnormal results. Alcohol consumption was found in 30.6% of the patients. uEtG was found to be the strongest marker of alcohol consumption (odds ratio = 414.5, P < 0.001) and provided a more accurate prediction rate of alcohol consumption [area under receiving operating characteristic (ROC) curve = 0.94] than CDT (area under ROC curve = 0.63, P < 0.001) and AUDIT-c (area under ROC curve = 0.73, P < 0.001). The combination of uEtG and AUDIT-c showed higher accuracy in detecting alcohol consumption in comparison with the combination of CDT and AUDIT-c (area under ROC curve = 0.98 versus 0.80, P < 0.001). Furthermore, uEtG was the most useful marker for detecting alcohol consumption in patients with negative AUDIT-c results. In conclusion, the combination of AUDIT-c and uEtG improves the detection of alcohol consumption in LTCs and LTRs. Therefore, they should be used routinely for these patients. Liver Transpl 20:815–822, 2014. © 2014 AASLD.
Provide feedback or get help You are viewing our new enhanced HTML article.
If you can't find a tool you're looking for, please click the link at the top of the page to "Go to old article view". Alternatively, view our Knowledge Base articles for additional help. Your feedback is important to us, so please let us know if you have comments or ideas for improvement.
Alcohol Use Disorders Identification Test
Alcohol Use Disorders Identification Test for Alcohol Consumption
liver transplant candidate
liver transplant recipient
mean corpuscular volume
negative predictive value
positive predictive value
receiving operating characteristic
urinary ethyl glucuronide
oxidazed nicotinamide adenine dinucleotide
reduced nicotinamide adenine dinucleotide
Alcoholic liver disease is one of the most common indications for liver transplantation (LT) in both Europe and the United States.[1, 2] Alcohol relapse after LT has been reported in 11% to 49% of patients,[3-5] and it is associated with graft dysfunction and a worse survival rate in comparison with abstinent patients.[6-8] Therefore, an assessment of drinking status is crucial to the selection of liver transplant candidates (LTCs). Most transplant centers in the United States require a 6-month period of abstinence before a patient is considered eligible for LT because it has been shown to predict a low rate of relapse into drinking after LT.[9, 10]
However, after a patient is considered eligible for LT by a health care professional experienced in the management of patients with addictive behavior, the transplant hepatologist needs to check the drinking status. Verbal reports by LTCs may underestimate their actual drinking habits, particularly if they believe that it may delay or prevent LT. Consequently, a valid screening tool is needed for the assessment of drinking status during the pretransplant period.
In the general population, markers of damage such as gamma-glutamyl transpeptidase (GGT), the aminotransferase (AST)/alanine aminotransferase (ALT) ratio, and the mean corpuscular volume (MCV) have been shown to be useful in detecting alcohol consumption.[11, 12] However, in patients with liver disease and/or liver transplant recipients (LTRs), their accuracy decreases.
Carbohydrate-deficient transferrin (CDT) is another marker of alcohol consumption that can be used to identify continuous, excessive drinking (50-80 g of ethanol/day) over a period of at least 1 week. A limitation of CDT is that it is unable to identify lower alcohol consumption, and its accuracy is limited in patients with liver cirrhosis. This reduced accuracy is due at least in part to methodological factors. A conventional immunoassay may produce several false-positive results because liver disease and perhaps other factors such as di-tri bridging can lead to falsely elevated results. As a result, it has been stated that CDT should be measured with high-performance liquid chromatography, particularly in patients with chronic liver disease.
Other direct markers of alcohol consumption such as serum ethanol (sEtOH) and urinary ethanol (uEtOH) can be used to identify small amounts of alcohol, but their application is limited by the time of detection, which ranges from 8 hours for sEtOH to 24 hours for uEtOH.
Recently, urinary ethyl glucuronide (uEtG) has been shown to be a reliable screening tool for the detection of alcohol consumption in LTCs and LTRs.[13, 17] uEtG is detectable after the consumption of very small amounts of ethanol (<10 g), and it can be found in the urine for up to 80 hours after the last drink. Ethyl glucuronide (EtG) can be measured by liquid chromatography–mass spectroscopy and liquid chromatography–tandem mass spectroscopy. A faster and less costly method involving an immunoassay has been developed and validated. Recently, data have shown that uEtG is capable of improving the detection of alcohol consumption in LTCs and LTRs. Furthermore, the determination of EtG in hair has been shown to improve the evaluation of long-term alcohol abstinence in LTCs.
As for the self-reporting of alcohol consumption, it is well known that verbal reports of drinking behavior by LTCs may underestimate their actual drinking. Consequently, it is important to use a validated screening test for the detection of alcohol consumption. To date, only the Timeline Followback (TLFB) test has been used for LTCs and LTRs. The Alcohol Use Disorders Identification Test for Alcohol Consumption (AUDIT-c) represents a simple and reliable test for the detection of alcohol consumption and hazardous drinking and has been proposed by the World Health Organization for the assessment of hazardous drinking in primary care. The Italian version has also been validated. Nevertheless, it has never been applied to LTCs and LTRs.
The aim of our study was to evaluate the accuracy of AUDIT-c, CDT, uEtG, sEtOH, uEtOH, and other indirect markers in the detection of alcohol consumption in LTCs and LTRs and thereby assess the best screening tools for these patients.
PATIENTS AND METHODS
From January 2011 to December 2012, 121 patients—LTCs affected by alcoholic cirrhosis and LTRs with alcoholic cirrhosis before LT—were consecutively enrolled and prospectively evaluated as outpatients in the care management program of our unit. The diagnosis of cirrhosis was based on histological findings (when available) or a combination of clinical, biochemical, ultrasonographic, and endoscopic findings. The study was approved by the ethics committee of the University of Padua and the General Hospital of Padua (protocol identification number 2068P). Written informed consent was obtained from each patient.
In the pretransplant setting, patients were checked with an individualized frequency ranging from every week to every month according to the severity of their liver disease. In the posttransplant setting, the frequency of scheduled visits was dependent on the time since LT and/or the development of complications. Accordingly, LTRs without complications were checked every month during the first 6 months after LT, every 2 to 3 months during 6-12 months after LT, and every 3 to 6 months thereafter. A relapse into drinking by LTRs was considered to be a medical complication. LTRs with a medical complication were checked more closely according to the type, severity, and evolution of the complication. The individualized frequencies of visits for these LTRs ranged from every week to every month.
For the purpose of this study, patients were interviewed during each visit by a hepatologist with respect to recent alcohol consumption; AUDIT-c was administered by the physician, and the score was calculated on the basis of responses to the following 3 questions with all types of alcohol combined: (1) the overall frequency of drinking, (2) the usual quantity of drinks consumed, and (3) the frequency of the consumption of 5 or more drinks. AUDIT-c was preferred to other tools such as the full Alcohol Use Disorders Identification Test (AUDIT) and TLFB because its compilation is simpler and less time consuming and it is, therefore, more suitable for assessing outpatients. AUDIT-c focuses on alcohol consumption (frequency and amount of drinking and episodes of heavy drinking) rather than alcohol dependence and has been found to be as accurate as the full AUDIT for detecting heavy drinking and/or active abuse or dependence in the general population. During each visit, blood and urine samples were collected for the detection of all alcohol markers. Urine samples were taken under the supervision of a nurse waiting in front of the bathroom door. When alcohol consumption was detected via any of the alcohol markers and the patient denied alcohol consumption according to AUDIT-c, he or she was scheduled for another visit and confronted with the test results. LTCs and LTRs who reported alcohol consumption were referred to our alcohol addiction unit for further evaluation and treatment. For LTCs, a reassessment of their eligibility was performed on a case-by-case basis by an addictive behavior specialist.
The determination of EtG was performed with a homogeneous immunoenzymatic assay. The DRI EtG enzyme immunoassay (Microgenics Corp., Fremont, CA) is based on the competition between a drug labeled with glucose-6-phosphate dehydrogenase and free drug contained in the urine sample for a fixed numbered of specific antibody binding sites. Antigen-antibody binding modulates the enzymatic activity: in the absence of free drug in the urine specimen, the specific antibody binds the drug labeled with glycerol-3-phosphate dehydrogenase and causes a reduction in enzymatic activity. This phenomenon favors a direct relationship between the urinal drug concentration and the enzymatic activity. At a wavelength of 540 nm, the active enzyme converts oxidized nicotinamide adenine dinucleotide (NAD) into reduced nicotinamide adenine dinucleotide (NADH) with a consequent increase in absorbance that can be measured spectrophotometrically. The EtG assay was applied with the Cobas 6000 automatic analyzer (version c501, Roche Diagnostics). The determination of uEtG was performed on the day of the sample collection. A value > 500 ng/mL was considered positive.
The determination of ethyl alcohol in urine and plasma specimens was made with an enzymatic method (ETOH2, Roche Diagnostics). Alcohol dehydrogenase converts ethanol into acetaldehyde and reduces NAD+ into NADH plus H+. The NADH that is formed during the reaction (measured spectrophotometrically at 340 nm) is directly proportional to the ethanol concentration in a sample. The method was the same for both biological matrices, and it was applied with the same instrument used for the EtG assay.
The percentage quantification of CDT was carried out with an Agilent 1100 series high-performance liquid chromatography system with an ultraviolet detector (Agilent Technologies, United States). The separation of transferrin glycoforms was performed with an anion-exchange chromatography column (Bio-Rad Laboratories, Germany). A value higher than 2.1% was considered positive according to the manufacturer.
AST (cutoff < 45 U/L for men and < 35 U/L for women), ALT (cutoff < 50 U/L for men and < 35 U/L for women), GGT (cutoff < 65 U/L for men and < 45 U/L for women), and MCV (cutoff < 96 fL) were also assessed at each visit. The AST/ALT ratio was calculated. Values ≥ 2 were considered to be indicative of advanced alcoholic liver disease.
In order to identify any intentional dilution of urine samples, the urinary creatinine level was determined with the Jaffe reaction method.
Definition of Alcohol Consumption
The detection of alcohol consumption was based on patient self-report (AUDIT-c) or the acknowledgment by a patient who had previously denied it via AUDIT-c after he or she was informed of abnormal results for 1 or more markers of alcohol consumption (sEtOH, uEtOH, uEtG, and CDT).
A statistical analysis was conducted with SAS 9.2 for Windows (SAS Institute, Inc., Cary, NC). Quantitative variables are reported as means and standard deviations. Categorical variables are reported as counts and percentages in each category. On the basis of patients' self-reports after confrontation with their results, the sensitivity, specificity, positive predictive values (PPVs), negative predictive values (NPVs), positive likelihood ratios, negative likelihood ratios, and accuracy of the tests were calculated. Markers predicting alcohol consumption were identified separately with a logistic regression analysis for each predictor. Subsequently, all predictors that had a P value < 0.1 were included in a multivariate logistic regression analysis. Results are presented as P values with odds ratios (ORs) and 95% confidence intervals (CIs). Spearman's rank correlation was used to detect the correlation between alcohol markers and the amount of alcohol consumption. The comparison of markers of alcohol consumption in the detection of alcohol consumption was performed with the C statistic and the area under the receiving operating characteristic (ROC) curve.
Characteristics of Patients
Baseline demographic, clinical, and laboratory data for the patients included in this study are reported in Table 1. The mean age of the patients was 56.5 ± 9.2 years, and 94 (77.7%) were male. Hepatitis B virus and hepatitis C virus were detected in 6 (5%) and 23 patients (19%), respectively. Ninety-eight patients (81%) were LTCs, whereas 23 (19%) were LTRs. The mean age of the LTCs was 55.7 years, and 76 (77.6%) were male. The indications for LT in the LTCs were decompensated liver cirrhosis (n = 77) and hepatocellular carcinoma (n = 21). As for the LTRs, the mean age was 60.0 years old, and 18 (78.3%) were male. The indications for LT in the LTRs were decompensated liver cirrhosis (n = 13), hepatorenal syndrome (n = 3), and hepatocellular carcinoma (n = 7).
Table 1. Demographic and Clinical Characteristics of the Patients
|Age (years)a||56.5 ± 9.2||55.7 ± 8.5||60.0 ± 11.2|
|Sex [n (%)]|| || || |
|Male||94 (77.7)||76 (77.6)||18 (78.3)|
|Female||27 (22.3)||22 (22.4)||5 (21.7)|
|Ethnicity (n)|| || || |
|Associated viral etiology of liver disease [n (%)]||29 (24)||22 (22.4)||7 (30.4)|
|Alcohol-dependent patients [n (%)]b||39 (32.2)||33 (33.7)||6 (26.1)|
|Alcohol abusers [n (%)]b||82 (67.8)||65 (66.3)||17 (73.9)|
|Alcohol consumption [n (%)]|| || || |
|Yes||37 (30.6)||32 (32.7)||5 (21.7)|
|No||84 (69.4)||66 (67.3)||18 (78.3)|
|Alcohol consumption of drinkers (drinks/week)ac||14.1 ± 8.6||14.2 ± 8.9||13.6 ± 6.2|
Detection of Alcohol Consumption
Alcohol consumption was detected in 37 of the 121 patients (30.6%). Alcohol use was found in 5 of the 23 LTRs (21.7%) and in 32 of the 98 LTCs (32.7%). Seventeen (14%) reported alcohol consumption according to AUDIT-c. Twenty patients (16.5%) who denied any alcohol intake on AUDIT-c then reported alcohol consumption when they were confronted with abnormal results for at least 1 marker. Nineteen of these 20 patients (95%) were found to be positive for uEtG, and 5 (25%) were found to be positive for CDT (P < 0.0001). Only 1 of the 5 patients who were positive for CDT was negative for uEtG.
Correlation Between Alcohol Markers and Amount of Alcohol Consumption
Patients whose alcohol consumption was detected reported a mean consumption of 14.1 ± 8.6 drinks per week. We found a statistically significant correlation between the amount of weekly alcohol intake and uEtG [correlation coefficient (ρ) = 0.53, P < 0.001], AUDIT-c (ρ = 0.43, P < 0.001), CDT (ρ = 0.35, P < 0.01), and GGT (ρ = 0.35, P < 0.001; Table 2). No correlation was found between the amount of weekly alcohol intake and the AST/ALT ratio or MCV.
Table 2. Correlation Between Alcohol Markers and the Amount of Alcohol Consumption
Comparison of Alcohol Markers in the Detection of Alcohol Consumption
In Table 3, we report the sensitivity, specificity, PPVs, NPVs, positive likelihood ratios, negative likelihood ratios, and accuracy of alcohol markers. uEtG showed the highest sensitivity (89.2%; 95% CI = 79.2%-99.2%) and accuracy (95.9%; 95% CI = 92.2%-99.5%) in comparison with the other alcohol markers. Only 1 of the 34 patients who were found to be positive for uEtG denied alcohol consumption when she was confronted with her results. CDT performed better for LTRs (sensitivity = 60%, specificity = 100%, PPV = 100%, NPV = 90%, accuracy = 91.3%) versus LTCs (sensitivity = 25%, specificity = 95.5%, PPV = 72.7%, NPV= 72.4, accuracy = 72.5%). It is worth noting that for 9 LTCs, the correct quantification of CDT was not possible because they showed poor chromatographic resolution of disialotransferrin from trisialotransferrin (a phenomenon known as di-tri bridging). Among these LTCs, 3 were found to be actively consuming alcohol.
Table 3. Comparison of Markers of Alcohol Abuse in the Detection of Alcohol Consumption
|uEtG||89.2 (79.2-99.2)||98.8 (96.5-100)||97.1 (91.4-100)||95.4 (91.0-99.8)||74.92 (10.64-527.37)||0.11 (0.04-0.28)||95.9 (92.2-99.5)|
|AUDIT-c||45.9 (36.9-55.2)||100 (96.2-100)||100 (96.2-100)||80.8 (72.4-87.1)||—||0.54 (0.40-0.73)||83.5 (76.2-90.7)|
|CDT||29.7 (15.0-44.5)||96.4 (92.4-100)||78.6 (57.1-100)||75.5 (67.3-83.7)||8.23 (2.44-27.76)||0.73 (0.59-0.90)||75.8 (67.0-84.6)|
|uEtOH||10.8 (0.8-20.8)||100 (100-100)||100 (100-100)||71.8 (63.6-80.0)||—||0.89 (0.80-1.00)||69.4 (59.6-79.3)|
|sEtOH||10.8 (0.8-20.8)||100 (100-100)||100 (100-100)||71.8 (63.6-80.0)||—||0.89 (0.80-1.00)||69.4 (59.6-79.3)|
|AST/ALT ratio > 2||29.7 (21.9-38.8)||81 (72.6-87.3)||40.7 (32.0-50.1)||72.3 (63.3-79.9)||1.56 (0.8-3.03)||0.87 (0.69-1.1)||65.3 (54.8-75.8)|
|GGT (U/L)||73.0 (58.7-87.3)||66.7 (56.6-76.8)||49.1 (35.9-62.3)||84.9 (76.2-93.5)||2.19 (1.53-3.14)||0.41 (0.23-0.70)||68.6 (58.6-78.6)|
|MCV (fL)||37.8 (22.2-53.5)||77.4 (68.4-86.3)||42.4 (25.6-59.3)||73.9 (64.7-83.0)||1.67 (0.94-2.96)||0.80 (0.61-1.06)||65.3 (54.8-75.8)|
In the univariate analysis, AUDIT-c (OR = 144.26; 95% CI = 7.68 to >999.99), uEtG (OR = 414.53; 95% CI = 61.11 to >999.99), CDT (OR = 9.98; 95% CI = 2.68-37.16), and GGT (OR = 5.10; 95% CI = 2.18-11.93) were found to be predictors of alcohol consumption (Table 4). In the multivariate analysis, AUDIT-c (OR = 274.25; 95% CI = 7.15 to >999.99) and uEtG (OR = 493.81; 95% CI = 51.29 to >999.99) were found to be the only independent predictors of alcohol consumption (Table 5). Subsequently, the accuracies of alcohol markers (uEtG, AUDIT-c, and CDT) were compared. uEtG (C statistic = 0.94; 95% CI = 0.89-0.99) was more accurate than AUDIT-c (C statistic = 0.73; 95% CI = 0.65-0.81, P < 0.001), CDT (C statistic = 0.63; 95% CI = 0.55-0.71, P < 0.001), and the combination of AUDIT-c and CDT (C statistic = 0.80; 95% CI = 0.72-0.88, P = 0.007; Tables 6 and 7). Furthermore, the combination of AUDIT-c and uEtG (C statistic = 0.98; 95% CI = 0.96-1.00) was found to slightly improve the accuracy of uEtG (P = 0.057; Table 7) and to be the most accurate combination of alcohol markers. The addition of CDT to AUDIT-c and uEtG was unable to further improve the model.
Table 4. Prediction of Alcohol Consumption With Alcohol Markers According to the Univariate Analysis
|AUDIT-c||144.26||7.68 to >999.99||<0.001|
|uEtG||414.53||61.11 to >999.99||<0.001|
Table 5. Risk Estimation of Alcohol Consumption With Alcohol Markers According to the Multivariate Analysis
|AUDIT-c||274.25||7.15 to >999.99||0.003|
|uEtG||493.81||51.29 to >999.99||<0.001|
Table 6. Diagnostic Accuracy of Markers of Alcohol Consumption
|AUDIT-c + CDT||0.80||0.72-0.88|
|AUDIT-c + uEtG||0.98||0.95-1.00|
|AUDIT-c + CDT + uEtG||0.98||0.96-1.00|
Table 7. Comparison of the Diagnostic Accuracy of Markers of Alcohol Consumption
|uEtG versus AUDIT-c||0.21||0.11-0.32||<0.001|
|uEtG versus CDT||0.31||0.22-0.40||<0.001|
|uEtG versus AUDIT-c + CDT||0.14||0.04-0.24||0.007|
|AUDIT-c + uEtG versus uEtG||0.04||−0.01 to 0.09||0.057|
|AUDIT-c + uEtG versus CDT + uEtG||0.18||0.09-0.27||<0.001|
|AUDIT-c + CDT + uEtG versus AUDIT-c + uEtG||0.002||−0.01 to 0.01||NS|
Although the length of abstinence from alcohol consumption is a matter of debate for patients who are potentially candidates for LT, there is no doubt that abstinence is required in the selection process for LT in patients with alcohol-related cirrhosis. A period of abstinence before LT is one of the criteria necessary to limit relapse into alcohol consumption after LT. There is no doubt that a small proportion of patients undergoing transplantation (LTRs) return to damaging drinking behavior. This can result in graft loss or patient death because of noncompliance with immunosuppressive therapy and/or direct hepatotoxic effects of alcohol on the graft.[6-8] Therefore, the monitoring of abstinence from alcohol consumption is a critical step in the management of LTCs and LTRs. The results of our study show that uEtG is the most accurate marker of alcohol consumption in this clinical setting. In our study, uEtG was capable of improving the detection of alcohol consumption in comparison with all other conventionally used markers and patient self-reporting. The enhanced accuracy of uEtG in comparison with indirect markers in the detection of alcohol consumption has already been previously reported[13, 17]; therefore, the results of this study were almost to be expected. On the other hand, the higher accuracy of uEtG versus AUDIT-c and CDT—the most original result of this study—deserves some comment. As far as AUDIT-c is concerned, it should be highlighted that it has not been previously reported for the detection of alcohol consumption in LTCs and LTRs. AUDIT-c was preferred to other tools such as TLFB for collecting self-reported drinking information because its compilation is simpler and less time-consuming and it is, therefore, more suitable for assessing outpatients. AUDIT-c was found to be an independent predictor of alcohol consumption and, despite its low sensitivity, was capable of detecting alcohol consumption in 3 patients who were found to have normal values for all other markers of alcohol consumption. The accuracy of AUDIT-c for the detection of alcohol consumption in our cohort of patients was lower than that reported in previous studies for the general population. This result is not surprising because it is well known that LTCs and LTRs are used to denying alcohol consumption, especially when verbal self-reporting is adopted. This is particularly true for LTCs, who know that an admission of alcohol consumption can prevent or at least delay their eligibility for LT.
As for CDT, another relevant result of the study shows that it is not an independent predictor of alcohol consumption in LTCs and LTRs (Table 4). Only in 5 of 20 patients who had denied alcohol consumption on AUDIT-c did an abnormal increase in CDT contribute to the detection of alcohol consumption by the patient, and in only 1 patient (an LTR) was it not associated with an abnormal value of uEtG. In good agreement with previous studies,[27, 28] CDT showed lower accuracy in LTCs versus LTRs in our study. This is not surprising because CDT loses its accuracy in patients with liver diseases and particularly in patients with liver cirrhosis.[4, 27, 29] The low accuracy of CDT in patients with cirrhosis is related first to a methodological factor. It has been suggested that an elevated value of CDT, detected by means of a conventional immunoassay, does not accurately predict alcohol consumption in patients with advanced liver disease because the immunoassay detects several false-positive results. The false-positive results are related to the fact that liver disease and perhaps other factors such as di-tri bridging can lead to falsely elevated results. Although the molecular basis for this effect is not known, the relationship between di-tri bridging and chronic liver disease is so strong that it points to di-tri bridging being “suggestive of liver disease.” As a result, it has been stated that CDT should be measured with high-performance liquid chromatography, particularly in patients with chronic liver disease. In fact, high-performance liquid chromatography can avoid the negative effects of liver abnormalities on the chromatographic separation of transferrin glycoforms by minimizing the reporting of false positives but without providing for them a valid detection of alcohol consumption. In our study, the use of high-performance liquid chromatography was not able to improve the accuracy of CDT in patients with chronic liver disease. In good agreement with these observations, the accuracy of CDT was low for the entire series of patients in our study. CDT seemed to recover accuracy to some extent when we moved from LTCs to LTRs, although the small number of LTRs in our study did not allow reliable conclusions regarding the accuracy of CDT in these patients.
The first strong point in favor of the use of uEtG for monitoring abstinence is that its value is not affected at all by liver disease. Clinical studies have shown that uEtG has similar accuracy in patients with advanced liver disease versus the general population.[18, 30] The low accuracy of CDT was also related to the particular condition of the patients who were included in our study. These patients were closely followed because either they were in the selection process for LT or on the waiting list for LT or they had already undergone LT. It can be easily assumed that for LTCs, not only the fear of causing further liver damage but also the fear of being excluded from the transplant waiting list may have limited the extent of alcohol consumption. Similarly, in the LTRs, the complexity of the procedure to which they had been subjected and the fear of breaking a moral deal with the transplant team may have produced the same effect. What is certain is that the alcohol consumption reported by our patients is far from that required to achieve a CDT value higher than 2.1%. Therefore, the second strong point in favor of the use of uEtG for monitoring abstinence in LTCs and LTRs is its higher sensitivity in comparison with CDT for these patterns of alcohol consumption. The consumption of even small amounts of ethanol (≈10 g) can result in positive uEtG findings for up to 80 hours after the last drink.
As for the last point, it is important to consider that the detection of a small amount of alcohol consumption can have different impacts on the management of LTCs and LTRs. In patients with cirrhosis, even a small amount of alcohol is able to significantly increase the portal pressure, and it can increase the risk of gastrointestinal bleeding and other complications. As a result, the prompt identification and treatment of active drinkers is mandatory. In this setting, because of its high sensitivity, uEtG represents a useful tool for detecting active drinkers and thereby allowing their prompt referral to intensive and multidisciplinary management, as mentioned previously.
In LTRs, different patterns of alcohol consumption have been shown. In clinical practice, it is important to differentiate between slips (the resumption of drinking small amounts) and relapse (a return to addictive drinking). Among LTRs, the resumption of addictive drinking is associated with injury to the graft and mortality.[6-8] In this setting, an isolated uEtG-positive result can reveal a slip. Although slips are not necessarily associated with graft dysfunction, they may be the harbinger of a relapse. Consequently, it is very important to identify slips in order to enable the patient to get help and, therefore, prevent a relapse.
The last point in favor of uEtG versus CDT is related to the fact that uEtG in our country is somewhat cheaper (approximately $7/test) than CDT (approximately $30/test). We realize that costs may vary widely across countries; nevertheless, at least the cost of CDT is the same as that in the United States.
Our study has some limitations. First of all, we used an enzyme immunoassay for the measurement of uEtG, and the results were not confirmed with liquid chromatography–mass spectroscopy or liquid chromatography–tandem mass spectroscopy. Nevertheless, the DRI EtG enzyme immunoassay has been validated in different studies.[13, 19] Moreover, it is worth noting that the ethanol contained in mouthwash solutions, alcohol-based hand sanitizers, and the ingestion of baker's yeast can lead to false-positive results,[35, 36] although the level of uEtG is usually expected to remain below the cutoff of 500 ng/mL. However, these possibilities should always be taken into account when one is interpreting test results, and they are the reason that a patient was considered to be a true positive for alcohol consumption only after his or her own admission. As for this last point, we would like to emphasize that the assessment of alcohol consumption with the use of biomarkers goes in parallel with the promotion of mutual trust between physicians and patients.
In conclusion, the combination of AUDIT-c and uEtG substantially improves the detection of alcohol consumption in these patients. We suggest using uEtG and AUDIT-c at each visit for LTCs and randomly for LTRs. However, uEtG and AUDIT-c should be used at each visit even for LTRs whenever clinical suspicion or risk factors for relapse into drinking occur.