Therapeutic plasma exchange in ACLF: what the randomised trial found
Swaroop and colleagues randomised 194 adults with ACLF to therapeutic plasma exchange plus standard medical therapy or standard therapy alone. This Evidence Deep Dive examines the study population, treatment protocol, 28- and 90-day outcomes, and the questions raised by the difference between early and later survival.
Why this study matters
ACLF carries a high risk of early death, while liver transplantation remains the only definitive treatment for suitable patients. An intervention that stabilises organ failure could create time for native-liver recovery or transplantation.
This trial moves the evidence for plasma exchange in ACLF beyond largely observational and heterogeneous studies into a randomised evaluation with a patient-centred primary outcome. It also tests a complex, resource-intensive intervention in a high-risk population. Understanding exactly who was enrolled, how treatment was delivered and how the survival difference changed over time is central to interpreting what the study adds.
Study at a glance
- Design: Single-centre, open-label, parallel-group randomised controlled trial
- Population: 194 adults aged 18–60 with EASL-CLIF ACLF grades 1–3b
- Intervention: Five consecutive daily sessions of therapeutic plasma exchange plus standard medical therapy
- Comparator: Standard medical therapy alone
- Primary outcome: All-cause mortality at 28 days
- Registration: CTRI/2022/01/039094
How the study was conducted
Between February 2022 and March 2025, the investigators randomly assigned 194 patients equally to therapeutic plasma exchange plus standard medical therapy or standard medical therapy alone. Randomisation used permuted blocks with reported allocation concealment. Blinding was not feasible for this intervention, but mortality was an objective primary outcome and the principal analysis was by intention to treat.
The plasma-exchange protocol specified five consecutive daily sessions. Each session exchanged approximately 4.1 litres of plasma and replaced approximately 3.5 litres using fresh frozen plasma and albumin.
Important exclusions included:
- active bacterial or fungal infection;
- haemodynamic instability despite fluids and vasopressors;
- significant comorbidity;
- hepatocellular carcinoma;
- previous liver transplantation;
- pregnancy; and
- death within 48 hours of admission.
The enrolled population was young, with a mean age of approximately 39 years. About two-thirds had alcohol-related liver disease, 64.5% had ACLF grade 2 and the median patient had two organ failures.
- Study design
- Single-centre, open-label, parallel-group randomised controlled trial with 1:1 allocation and intention-to-treat analysis.
- Population
- 194 adults aged 18–60 with EASL-CLIF-defined ACLF grades 1–3b. The population was young (mean age approximately 39 years), about two-thirds had alcohol-related liver disease, 64.5% had ACLF grade 2 and the median patient had two organ failures.
- Intervention or exposure
- Five consecutive daily sessions of therapeutic plasma exchange in addition to standard medical therapy. Each session exchanged approximately 4.1 litres of plasma and replaced approximately 3.5 litres, using fresh frozen plasma and albumin.
- Comparator
- Standard medical therapy alone.
What the study found
- 28-day mortality: 43/97 (44.3%) with plasma exchange versus 62/97 (63.9%) with standard therapy. The absolute risk reduction was 19.6% (95% CI 5.8%–33.3%; p=0.006). The calculated relative risk was approximately 0.69 (95% CI 0.53–0.91), corresponding to a number needed to treat of about five.
- 90-day mortality: 64/97 (66.0%) versus 67/97 (69.1%; p=0.64).
- Treatment delivery: 72/97 patients received at least three exchanges and 56/97 (57.7%) completed all five. Haemodynamic instability was the commonest reason for stopping.
- New infection: 21.7% versus 11.8% (p=0.062).
- Biochemistry and organ failure: Plasma exchange produced greater early reductions in ammonia, bilirubin and INR, with more apparent resolution of hepatic and coagulation failure by day 7.
The prespecified 28-day outcome favoured plasma exchange, while the groups had similar mortality by day 90. The study therefore describes a substantial early survival difference that was not sustained at the later time point.
Interpreting the findings
Features that support interpretation
- Random allocation provides a stronger basis for comparing the intervention with standard therapy than the earlier observational literature.
- The primary outcome was objective, clinically important and analysed according to randomisation.
- Allocation concealment was reported, and baseline characteristics were well balanced.
- Treatment completion and reasons for discontinuation were reported transparently, allowing feasibility to be considered alongside the outcome data.
- Follow-up to 90 days shows how the between-group difference changed beyond the primary endpoint.
- The high control-group mortality confirms that the trial enrolled a population at substantial short-term risk.
Context and remaining uncertainty
Duration of the observed benefit
The trial met its 28-day primary endpoint, but mortality was similar between groups at 90 days. Further studies will need to establish whether an early survival window can translate into native-liver recovery or successful transplantation.
Population studied
Eligibility was limited to adults aged 18–60 without active infection or persistent haemodynamic instability. Many patients seen in liver critical care have infection, vasoplegia or escalating organ-support requirements, so the clearest inferences apply to the selected population enrolled.
Interpretation of biochemical outcomes
Plasma exchange directly removes bilirubin and ammonia, while fresh frozen plasma replaces coagulation factors and lowers INR. These changes demonstrate the biological effect of the procedure, but do not by themselves distinguish hepatic regeneration from extracorporeal clearance and replacement.
Treatment delivery
Only 56 of 97 patients completed all five intended sessions, with haemodynamic instability the commonest reason for stopping. This makes treatment tolerance, dose, stopping rules and patient selection important parts of any future evaluation.
Safety and setting
New infections were numerically more frequent after plasma exchange, although the trial was not powered for a definitive safety comparison. The study was also single-centre and open-label. Larger multicentre studies could examine infection, vascular-access complications, electrolyte disturbance, plasma exposure, co-interventions and resource use across different aetiologies and transplant systems.
Relevance to liver critical care
For liver critical care, the study provides randomised evidence that therapeutic plasma exchange can alter early outcomes in selected, haemodynamically stable adults with ACLF. The 28-day mortality difference is clinically important, but the absence of a 90-day difference means the mechanism and purpose of that early survival window require further definition.
A key clinical question is whether early stabilisation can function as a bridge to native-liver recovery or transplantation. The treatment protocol was also demanding, with just over half of participants completing all five exchanges, so feasibility and tolerance are integral to interpreting its possible role.
Questions that remain
- Which clinical or biological features identify patients most likely to benefit?
- Can an early survival window be converted into improved 90-day transplant-free survival?
- How does access to liver transplantation affect the longer-term outcome?
- What is the most effective number, volume and timing of exchanges?
- Which stopping rules best balance possible benefit, treatment tolerance and resource use?
- How should infection, haemodynamic instability and vasopressor-dependent disease affect eligibility?
- Are the findings reproducible across multiple centres, different ACLF aetiologies and different transplant systems?
- What are the effects on infection, vascular-access complications, electrolyte disturbance, plasma exposure and cost?
Full reference and source
Swaroop S, Biswas S, Aggarwal A, et al. Therapeutic plasma exchange improves short-term survival in patients with acute-on-chronic liver failure: a randomized controlled trial. Hepatology. Published online March 30, 2026. doi:10.1097/HEP.0000000000001755.