Liver Support Systems Show No Clear Survival Benefit in Acute Liver Failure

Acute liver failure (ALF) is a rapidly progressive, life-threatening condition characterized by severe hepatocyte injury, coagulopathy, and hepatic encephalopathy in patients without pre-existing liver disease. Without liver transplantation or spontaneous recovery, mortality remains extremely high. Liver support systems — both artificial (cell-free, detoxification-based) and bioartificial (incorporating living hepatocytes) — have been developed as bridge therapies to sustain patients until transplantation or hepatic regeneration. Despite decades of use and ongoing clinical interest, their true efficacy has remained contested. This Cochrane review, updated through July 2025, provides the most comprehensive synthesis to date.

The review included 11 parallel-group randomized clinical trials (RCTs) enrolling 681 participants (342 experimental, 339 control), conducted across Europe, India, and the USA. Eight trials evaluated artificial liver support systems and three evaluated bioartificial systems. All participants were treated in intensive care units. The age range was 10–69 years, with a small number of pediatric patients whose data could not be separated. Trials were published between 1973 and 2019, and two were available only as abstracts. The primary outcomes were all-cause mortality at day 28 and at maximal follow-up (up to 90 days for mortality, 360 days for transplantation and multi-organ failure).

For 28-day all-cause mortality, pooled analysis of 9 studies (539 participants) yielded a risk ratio (RR) of 0.83 (95% CI 0.58–1.19), with substantial heterogeneity (I² = 66%) and very low certainty of evidence. At maximal follow-up, 11 studies (681 participants) showed RR 0.82 (95% CI 0.63–1.07; I² = 52%), again very low certainty. Neither result reached statistical significance. Serious adverse events showed a rate ratio of 0.93 (95% CI 0.81–1.07; I² = 0%; 11 studies, 681 participants), and liver transplantation rates were virtually identical between groups (RR 1.03, 95% CI 0.87–1.21; I² = 0%; 5 studies, 519 participants).

For hepatic encephalopathy, only 2 studies with 34 participants contributed data, yielding RR 0.40 (95% CI 0.16–1.04; I² = 0%) — a numerically promising but statistically non-significant and extremely uncertain result given the tiny sample. Multi-organ failure rate ratio was 0.92 (95% CI 0.72–1.18; I² = 0%; 6 studies, 501 participants), suggesting no meaningful difference. No trial reported health-related quality of life. All seven prespecified outcome results were downgraded to very low certainty using GRADE, primarily due to high risk of bias across all trials, imprecision from small sample sizes, and heterogeneity in the mortality analyses.

The clinical implications are sobering. Despite decades of development and ongoing clinical use, liver support systems have not demonstrated a clear survival or functional benefit in ALF in any adequately powered, well-designed trial. Most devices studied were produced by small manufacturers and are no longer commercially available; none of the bioartificial systems are currently on the market. The absence of ongoing RCTs is particularly concerning, as it suggests the field may be stagnating. For clinicians, these findings reinforce that liver support systems should not replace standard care or delay transplantation evaluation, and their use should ideally occur within a research framework. For patients and families, the evidence does not support confident expectations of benefit from these technologies in their current form.