Comparison of LDH activity and LDH-M to predict death in patients with acetaminophen-induced acute liver failure

TL;DR

This study compares LDH activity and LDH-M levels to predict death in patients with acetaminophen-induced liver failure.

Contribution

The study evaluates whether LDH-M provides better prognostic accuracy than total LDH in predicting mortality in acute liver failure.

Findings

Both total LDH and LDH-M levels were higher in non-survivors compared to survivors.

Total LDH performed similarly to or slightly better than LDH-M and MELD scores in predicting mortality.

MELD-LDH score showed modest improvement over the standard MELD score.

Abstract

Liver transplantation is often the only life-saving intervention in acute liver failure (ALF), so it is critical to identify all ALF patients who need a transplant to survive. However, transplantable organs are scarce and recipients face significant life-threatening risks, so it is also important to avoid transplantation when possible. Thus, prognostics that can predict death in ALF with both high sensitivity and specificity are needed. We previously demonstrated total lactate dehydrogenase (LDH) activity may be useful for this purpose. However, LDH is a tetramer of LDH-M and LDH-H, and proteomics indicated only the liver enzyme - LDH-M - increases in non-survivors. LDH-H decreased. Because total LDH comprises both, it is possible that LDH-M could have even better prognostic performance. To test that, we compared total LDH and LDH-M in a subset of samples from our prior study of survivors and non-survivors of acetaminophen (APAP)-induced ALF. The model for end-stage liver disease (MELD) score and the MELD-LDH score were also calculated. Both total LDH and LDH-M values were greater in non-survivors than survivors, but there was no significant difference in prognostic metrics between them. Overall, total LDH performed similarly to or modestly better than both LDH-M and the MELD. The MELD-LDH score also performed somewhat better than the MELD.

See also the graphical abstract(Fig. 1).

Figures