Monte Minute‎ > ‎Monte Minute‎ > ‎

Alcoholic Hepatitis and Aspects of Acid-Base Disturbances - 9/10/2012

posted Sep 10, 2012, 8:23 AM by Rohit Das   [ updated Dec 27, 2012, 6:56 AM by Purnema Madahar ]

A new week, a new daily. Today, I’d like to talk about a couple of things – alcoholic hepatitis, and some specific acid-base issues.

Alcoholic hepatitis is within a very large spectrum of alcoholic liver disease – a very nice review of this topic is attached. Let’s focus on some of the important characteristics of this very important consequence of excessive boozing:

  • Alcoholic liver disease (ALD) is a significant contributor to liver-related mortality, on the scope of HCV related liver disease. In parts of the world where inhabitants particularly like alcohol (France/Spain), mortality is around 30 deaths/100,000 persons per year. In the U.S., estimates from 2003 cite a mortality rate of 5-6 deaths/100,000 persons per year.
  • ALD exists within a large spectrum of disease – steatosis is characteristic of moderate, short-term alcohol use, whereas long-term use will lead to steatohepatitis, fibrosis, and potentially long-term, cirrhosis. Studies have shown that progression of liver disease is directly related to the absolute amount of alcohol intake. At 30g/day (around 2.5-3 standard drinks, more practically) over 10 years, estimates of incidence of cirrhosis is 5-7% at 5 years. Intake above that threshold leads to further climbing incidence.
  • Figure 2 of the review article gives a nice pictorial overview of the pathogenesis of ALD. Importantly, since the majority of heavy alcohol users DO NOT have progressive liver injury, it is important to realize that genetics, concurrent liver disease (NAFLD, viral hepatitis) and gender are all contributing factors to potential progression of ALD.

Alcoholic hepatitis is the acute form of ALD, presenting with hepatomegaly, low-grade fever, jaundice and anorexia. Complications of portal hypertension can also be present, which is not surprisingly associated with a worse prognosis.

  • Characteristically, for not very clear reasons, the AST/ALT ratio is > 2:1, which occurs in about 70-80% patients with ALD. Also, absolute transaminase levels are usually never above 500, and typically less than 300.
  • Severe alcoholic hepatitis has a grim prognosis, with a six month mortality rate of up to 50%. What does “severe” mean? Well there are several ways to assess for that, including the Maddrey’s score (>32 = bad), Glasgow hepatitis score (>9 = bad), and MELD (>21 = bad). All of these have been fairly well validated as predictive tools for severe disease. The attached review provides a nice summary.
  • Regarding treatment; steroids have been decently studied, and the data is supportive for the severe spectrum of disease. In one pooled analysis of randomized trials in 2002, among patients with encephalopathy and/or a Maddrey’s score > 32, those treated with steroids had significantly improved one-month survival. However, this survival benefit did not extend past one year, again reinforcing the overall poor prognosis of this disease.
  • Due to the presumed effects of TNF on the pathogenesis of alcoholic hepatitis, pentoxyfylline (which inhibits TNF synthesis) is also used, and has some data to go behind it (one month mortality of 25%, versus 45% in the placebo arm). The data suggests that the benefit may be specifically related to the prevention of HRS…which is interesting. However, pentoxyfylline, with steroids, does not seem to have an additive benefit in a preliminary randomized trial.
  • Above all else, though, is abstinence from alcohol. Even in later stages of liver disease, like cirrhosis, abstinence can lead to some degree of regression. Bottom line - abstinence prevents progression and improves survival.

Now, a couple of words on DKA and some interesting acid-base stuff to be aware of. DKA, as we all know (or hopefully know), is an anion-gap acidosis due to the accumulation of ketoacids in the setting of insulin deficiency. Couple of things to realize…

·  In patients with normal renal function, treatment of DKA leads to “ketouresis” (I made that word up). Ketoacids lost in the urine would have been metabolized to bicarbonate, and their excretion is referred to as loss of "potential bicarbonate." This effectively leads to a non-anion gap (hyperchloremic) acidosis. The presence of this DOES NOT mean that insulin can’t be switched to SQ – this is a self-limited issue and resolves with time

A bit about the mysterious Delta Anion Gap/Delta Bicarbonate…

· In utopia, where everything happens in the same compartment, excess acid is buffered by bicarbonate in a 1:1 ratio. Since we don’t live in utopia, and acid gets buffered in places outside the extracellular fluid (bones, intracellularly), the change in anion gap and bicarbonate are not exactly proportional.

·   Of the most common reasons for anion gap acidosis, lactic acidosis leads to a ratio of 1.6:1 (due to poor renal perfusion and impaired excretion of excess lactate) and DKA leads to a ratio usually around 1:1 (due to increased excretion of ketoacids, in patients with intact renal function).

·   This measurement is particularly useful in helping to diagnose mixed acid-base disturbances. For example, concurrent diarrhea will lead to a lower ratio (higher delta bicarb than expected), and a concurrent metabolic alkalosis/respiratory acidosis with compensation will lead to a higher ratio (lower delta bicarb than expected).

Hope this makes sense. I’ve attached a couple of decent review articles - one on DKA diagnosis/management, and another on interpretation of anion gap in the context of the delta/delta ratio.


 Watch out for the joint commission…but don’t let them stop you from reading!!!! Until tomorrow…

Cainelli, World J Hep 2012, Volume 4 (3): 81-90

Mathurin et. al., Journal of Hep 2002, Volume 36: 480-87

Diagnosis and Treatment of DKA and HHS
Chiasson et. al., CMAJ 2003, Volume 168 (7): 859-66

Reddy et. al., Int J of Clin Practice 2009, Volume 63: 1409-12