In patients with acute alcoholic hepatitis, does pentoxifylline reduce mortality?

Duncan Chambler

Pathological liver changes in relation to excessive alcohol can be classified into three types(1), all of which can overlap:
• Steatosis, or fatty liver
• Acute inflammatory hepatitis, or Alcoholic Hepatitis (AH)
• Cirrhosis

Fatty liver changes can be rapidly reversed with sustained abstinence, whilst cirrhotic leads to permanent fibrotic changes and loss of hepatic function. Alcoholic Hepatitis can be triggered with minimal ingestion alcohol, although significant drinking does not always produce hepatitic changes. Clinically, AH induces a SIRS response and is often the result of an acute physiological disturbance, such as gastrointestinal bleeding or infection. Management of AH is largely supportive, and can involve multi-organ support. Follow-up data from a large observational study(2) highlights the poor prognosis of this condition, especially if several organs show signs of acute injury or failure. A recent review(3) suggests that patients presenting with ALD related GI bleeding have a high mortality rate initially (48% unit mortality), but longer term survival is more likely than those presenting with encephalopathy or sepsis (mortality at 12 months: GI bleed 68%, encephalopathy 80%, sepsis 67–88%).

Searching Medline for “Alcoholic Hepatitis” and “Pentoxifylline”, as both MeSH headings and text words (title and abstract) reveals 65 articles (performed 15 May 2013), of which one is a Cochrane Review from 2009(4). This review identified only one fully published RCT(5), along with 4 abstracts. The above search revealed just one more recent relevant RCT(6) published since the Cochrane Review, although it appears to be the data from a previous abstract included in the Cochrane Review.

The Cochrane Review concluded that pentoxifylline use for AH could not be recommended. Although the main RCT(5) included in the analysis found a statistically significant benefit, drawing conclusions on one small RCT can lead to a type I (false positive) error. This RCT studied jaundiced patients, with a mDF≥32, clinical findings in keeping with hepatitis and a history of excessive alcohol consumption. The authors excluded patients with a gastrointestinal bleed, such as the case described here. The intervention group were administered 400mg pentoxifylline TDS, whilst the control group were given vitamin B (considered to be placebo). Recruitment, randomisation, allocation, maintenance and primary outcome measurement were all performed adequately. Mortality was 24% in the intervention arm, and 46% in the control arm (RR of death 0.59, 95% CI 0.35–0.97, p=0.037). This equates to an NNT of 4.5. Side effects included epigastric pain and gastrointestinal upset, causing 14% in the intervention arm to withdraw from treatment (compared to 2% in control arm).

Other studies have compared pentoxifylline with or against corticosteroids and drawn mixed conclusions. This aspect has not been considered for this expanded case summary.

---

References

1 Jackson P, and Gleeson D. Alcoholic liver disease. Continuing Education in Anaesthesia, Critical Care and Pain 2010; 10(3): 66-71.
2 Mackle IJ, Swann DG and Cook B. One year outcome of intensive care patients with decompensated alcoholic liver disease. British Journal of Anaesthesia 2006; 97: 496-8.
3 Flood S, Bodenham A, and Jackson P. Mortality of patients with alcoholic liver disease admitted to critical care: a systematic review. Journal of the Intensive Care Society 2012;
13(2): 130-5.
4 Whitfield K, et al. Pentoxifylline for alcoholic hepatitis. Cochrane Database of Systematic Reviews 2009; (4): 1361-6137; 1469-493X.
5 Akriviadis E, et al. Pentoxifylline improves short-term survival in severe acute alcoholic hepatitis: a double-blind, placebo-controlled trial. Gastroenterology 2000; 119(6): 1637-48.
6 Sidhu SS, et al. Pentoxifylline in Severe Alcoholic Hepatitis: A Prospective, Randomised Trial. Journal of the Association of Physicians of India 2012; 60: 20-2.