In patients with a first episode of severe alcoholic hepatitis non-responsive to medical therapy, early liver transplant increases 6-month survival
- Division of Gastroenterology and Hepatology, Department of Medicine, University of Wisconsin-Madison, Madison, Wisconsin, USA
- Correspondence to: Michael R Lucey
University of Wisconsin-Madison, 4245 UW Med Fndtn Centennial Bldg, 1685 Highland Ave Madison, WI 53705, USA;
In its most severe form, alcoholic hepatitis is associated with jaundice, coagulopathy, leucocytosis and rapid deterioration to death.1 When used in conjunction with abstinence from alcohol, corticosteroids improve 6-month survival for such severely affected patients.2 Treatment failures are identified after 7 days of therapy, using a simple mathematical instrument (http://www.lillemodel.com).3 Patients with alcoholic hepatitis who are unresponsive to medical management have a dire 6-month mortality.
Widely-used selection rules exclude patients with alcoholic hepatitis from liver transplantation (LT) unless they demonstrate 6 months of abstinence from alcohol.4 ,5 The study by Mathurin et al used early LT to examine the survival benefits of early LT among patients with severe alcoholic hepatitis with less than 6 months of alcohol abstinence.
This is a prospective non-randomised observational study in seven European centres of LT in severe alcoholic hepatitis. The inclusion criteria include: (1) first episode of alcoholic hepatitis, (2) severity (Maddrey's discriminant function >32), (3) failure to respond to prednisolone 40 mg/day for 7 days and rigorous medical treatment and (4) extensive psychological assessment. Patients had to have no psychiatric disorders and agree to lifelong alcohol abstinence and close supportive family members also had to agree.
Observed outcomes include: (1) 6-month survival from the day of starting corticosteroids, (2) the impact on the overall transplant patient profile and (3) abstinence after transplantation. Two historical control groups were constructed from the records of the largest recruiting centre (Lille, France).
The study cohort consisted of 26 patients. Of these, 24 were failures of treatment with corticosteroids. Alcoholic hepatitis was confirmed by biopsy in 23 subjects, and in all explants.6 Patients were placed on the transplant waiting list, on average, 13 days after corticosteroid failure. The median model for end-stage liver disease (MELD) at listing was 34. The 6-month survival was 77.8% in the transplanted group, compared with 23.8% in matched historical controls (p<0.001). Among the controls, 90% of deaths occurred within 2 months of failing corticosteroid therapy. Five of six deaths in the transplanted group were due to infection, with invasive aspergillus being the cause in four. This pilot program accounted for 2.9% of all transplants and 8.35% of transplants for alcoholic disease at the participating centres during the time of the study. Three patients returned to drinking after transplantation.
This pilot study strongly suggests that LT is life-saving for some patients with severe alcoholic hepatitis, who are unresponsive to medical management, and who have met stringent psychosocial requirements. Few patients returned to drinking after transplantation, and the proportion of transplants going to patients with alcoholic hepatitis was small.
Broader application of early transplantation would need a change of consensus by payers and regulatory agencies regarding the utility of the 6-month abstinence rule. The subjects in the present study benefited from rapid access to deceased donor organs, which might not be possible outside of a study. Conversely, since the listed patients had MELD scores in the mid-30s, it is likely that patients with severe alcoholic hepatitis would be favoured by a MELD-based allocation system.
Before early LT for alcoholic hepatitis receives broader application, several questions remain. In a repeat study, would a contemporaneous non-transplanted control group have the same poor outcome observed here in historical controls? How reproducible is the selection process described by Mathurin et al and is there a more simple way to achieve appropriate selection? Is restricting transplantation to patients with their first decompensating event appropriate? In view of the impact of post-transplant infection on mortality, would it be possible to achieve even better survival results with antifungal prophylaxis? Finally, would more widespread adoption of these selection criteria result in alcoholic hepatitis accounting for a greater proportion of total transplants? While alcoholic hepatitis remains a contraindication to LT, with this study, the door has been opened a crack.