Monitoring alcohol use on the liver transplant wait list: Therapeutic and practical issues

Authors

  • Andrea F. DiMartini,

    Corresponding author
    1. Departments of Psychiatry, University of Pittsburgh, Pittsburgh, PA
    2. Surgery, University of Pittsburgh, Pittsburgh, PA
    3. Starzl Transplant Institute, University of Pittsburgh Medical Center, Pittsburgh, PA
    • School of Medicine, University of Pittsburgh, 3811 O'Hara Street, Pittsburgh, PA 15213
    Search for more papers by this author
    • Telephone: 412-624-3373; FAX: 412-383-4846

  • Mary Amanda Dew

    1. Departments of Psychiatry, University of Pittsburgh, Pittsburgh, PA
    2. Psychology, University of Pittsburgh, Pittsburgh, PA
    3. Epidemiology, University of Pittsburgh, Pittsburgh, PA
    4. Biostatistics, School of Medicine, University of Pittsburgh, Pittsburgh, PA
    Search for more papers by this author

In this issue of Liver Transplantation, Hempel et al.1 report on the use of a specific serum biomarker, methanol, which is also present in alcoholic beverages, to detect recent alcohol consumption by patients with alcoholic cirrhosis who were beginning the evaluation process for liver transplantation (LT). In a side-by-side comparison, they found that 35% of patients were positive for methanol, whereas only 3% were identified by standard serum blood alcohol levels (BALs) or verbal reports. Although their study demonstrates that there are potentially better biomarkers than serum alcohol for detecting recent alcohol consumption, the need for improved methods of detecting covert alcohol use raises important issues for monitoring transplant candidates.

Even after patients are selected and entered onto LT wait lists, the rates of identified alcohol consumption are high. Up to 25% of wait-listed candidates use alcohol, with the majority of these discoveries made through the use of alcohol testing.2–4 Clearly, monitoring is required both during the evaluation phase and during the wait-list phase, but what is the best method? Although we found that our transplant clinical interviews after LT revealed the most episodes of alcohol use in comparison with other monitoring methods (including research assessments similar to the self-report measure used by Hempel et al.1), we surmised that this was due to the therapeutic relationship that we had established with the patients over the course of their transplantation.5 Our recent meta-analysis has also demonstrated that studies relying on self-reports of alcohol use after transplantation are equally as likely to identify alcohol use as studies using other methods.6 In contrast, in LT candidates, the discovery of such high rates of alcohol use by a biological marker when alcohol use has been denied during self-report assessments points to the very complex nature of these clinical interactions.

INTERVIEWS CANNOT BE THE SOLE METHOD OF MONITORING

In most populations of nontransplant patients seeking treatment for addictions, the identification of alcohol use is best and most frequently established through clinical interviews. Patients actively seeking treatment for addiction disorders voluntarily engage in therapeutic interactions specifically for the purpose of treatment, and the discussion of their current alcohol use is essential to their success. The answer to why we cannot expect such candidness from our candidates is rooted in the nature of the relationship. The pretransplant evaluation is not a therapeutic interaction but a judgment of the suitability of the individual for solid organ transplantation. Patients who are aware of this nuance may conceal aspects of their history or current behaviors that they feel may not be viewed positively. In the context of such an evaluation, clinicians also may consider the information provided by patients about their history and abstinence to be suspect. Even for patients voluntarily seeking addiction treatment, the psychological obstacles of shame, guilt, and denial often make discussions of their consumption patterns with treating clinicians difficult. These issues are heightened for patients seeking transplantation. This is especially true after they are wait-listed, and patients additionally recognize that the consequence of the disclosure or discovery of alcohol use could be removal from the wait list. It is notable that an older survey of LT programs found that 15% would permanently delist such individuals.7

Abbreviations:

BAL, blood alcohol level; LT, liver transplantation.

LT CANDIDATES DO NOT PERCEIVE A NEED FOR OVERSIGHT

Commonly, patients with alcoholic liver disease who are being considered for transplantation, rather than voluntarily choosing to stop drinking, have had the decision to stop drinking thrust upon them because of a specific health event, such as gastrointestinal bleeding or the diagnosis of cirrhosis. In particular, Weinrieb et al.2 found that LT candidates differed from other nontransplant patients with addictions in that they considered these other aspects of their health to be a priority over addiction treatment, were preoccupied with transplant management, and did not perceive a need for addiction counseling. Many did not consider themselves to have an addiction problem. Requests for addiction counseling and biological samples to monitor for alcohol use may be viewed as offensive by these patients. Weinrieb et al. suggested that the best approach to monitoring these patients would be outside the transplant clinic, where confidentiality could be maintained, and alcohol use, if disclosed, would not be revealed to the team except in life-threatening emergencies. Weinrieb et al. believed that this anonymity would increase a patient's willingness to disclose alcohol use, and this would ultimately enhance the patient's ability to seek and receive appropriate treatment for addictions.

WHICH BIOLOGICAL MARKERS SHOULD BE USED?

The choice of biological markers requires a consideration of the parameters and limits of biomarkers as well as their ease of use and acceptability. BALs have been the standard at most programs because of the widespread availability of this test. Positive screening results should be confirmed by gas chromatography and mass spectroscopy because of the possible risk of false positives.8 Breath analyzers have also been used for LT candidates to sample for expired alcohol in the breath,2 and they can approximate the concentration of alcohol in the blood. Because of its fast elimination, tools measuring alcohol will detect only very recent use. In this issue of the journal, Hempel et al.1 provide data suggesting the benefits of serum methanol levels versus BALs, but there are other biological markers to be considered. Although it is not widely used in routine clinical practice, urinary ethyl glucuronide, a conjugated minor ethanol metabolite with a longer detection window than ethanol, can be detected even after serum ethanol levels are negative. In a study of wait-listed LT candidates receiving alcohol rehabilitation treatment, ethyl glucuronide was significantly more sensitive in identifying alcohol use than self-reports or breath analyses.9 Although no patient endorsed alcohol use and only 1% of the breath analysis tests were positive, 25% of the ethyl glucuronide samples were positive, and this represented a striking 50% of the patients. Although the cohort was predominantly composed of high-risk individuals who were required to participate in addiction counseling as a precondition for LT listing, this study demonstrates the potential validity of various methods for determining alcohol use when the disincentives for reporting use are high. Carbohydrate-deficient transferrin is an abnormal form of a liver protein that is created from sustained heavy alcohol use. Carbohydrate-deficient transferrin may be detected in the blood even weeks after abstinence, but it can be elevated (and falsely suggest alcohol use) in patients with advanced liver disease.10 Hair toxicology analysis provides a longer window of ascertainment (90 days) and can identify the use of alcohol and other drugs. In Haller et al.'s study of transplant candidates,11 hair analysis indicated twice as many cases of use as other tests (self-reports, breath analysis, and urine analysis). Haller et al. suggested that, on the basis of the detection window for hair analysis, 2 negative tests would help to confirm 6 months of continuous abstinence. Finally, new technologies in law enforcement have demonstrated that continuous monitoring of alcohol use can be achieved by sweat analyses from a device worn next to the skin. Whether transplant programs will feel compelled to go to this extreme is uncertain.

RANDOM TESTING IS IMPORTANT

As Hempel et al.1 demonstrate in their study, random toxicology screens more often yield positive results than testing during a scheduled clinic appointment. In practice, patients would be notified about the need to provide a sample of blood or urine and given a limited time in which they could do this. For blood ethanol levels, such a random request requires the individual to go immediately for testing. With the longer duration of methanol elimination, Hempel et al. allowed patients up to 48 hours to complete the testing. Interestingly, more frequent random BAL testing was associated with fewer positives in 1 study,3 and the investigators interpreted this as possibly demonstrating that patients recognized their accountability to the LT team or even that BAL screening could be a form of intervention.

WHAT IS THE OPTIMAL ARRANGEMENT?

Regular monitoring while patients are on the wait list is critical for determining ongoing abstinence and providing treatment assistance when alcohol use is identified. Using a combination of methods (patient interviews, independent caregiver reports, and biochemical monitoring) provides the greatest yield because every method can add to the number of identified cases.2, 5 Regular interviewing and random sampling by a clinical team not connected to the transplant team may provide the most therapeutic arrangement and allow more candid disclosure. Whether LT teams would allow the results of testing and interviewing to remain confidential is debatable. Optimally, the development of a positive trustful relationship is needed so that a relapsing individual can be provided with appropriate help without fear of permanent removal from the transplant wait list. Finally, all patients, regardless of their reported history of alcohol or drug use, should be evaluated with toxicology screening at least during their initial transplant evaluation. This is a more equitable, less stigmatizing procedure than singling out those who have honestly reported their substance use history. Such universal testing would likely yield many more positive test results, and the transplant team should be prepared to deal with unexpected discoveries.

Ancillary