Navigating the therapeutic tightrope: Precision use of plasmapheresis and continuous renal replacement therapy in liver failure

Abstract

Liver failure, which includes acute liver failure (ALF) and acute-on-chronic liver failure (ACLF), is a life-threatening condition characterised by severe loss of hepatocytes, systemic inflammation, and multi-organ dysfunction, often leading to mortality rates exceeding 50%. Therapeutic plasma exchange (TPE, also known as plasmapheresis) and continuous renal replacement therapy (CRRT) are essential extracorporeal treatments that detoxify the blood, stabilise patients, and act as a bridge to recovery or transplantation. The debate surrounding the use of TPE and CRRT in liver failure (ALF and ACLF) is ongoing, with a recognised need to clearly define the patient groups, timing, and modality of application. This minireview synthesises evidence from 2016 to 2025, obtained from PubMed, EMBASE, and the Cochrane databases, and examines the mechanisms, indications, protocols, and outcomes of TPE and CRRT in ALF and ACLF. Correct patient selection, timing, and monitoring are essential to optimise benefits while minimising risks such as bleeding, infections, or citrate toxicity. This review examines various aspects of both therapies to help determine the most suitable treatment for liver failure. It highlights the importance of standardised guidelines to improve outcomes across different settings.

Key Words: Acute liver failure; Acute-on-chronic liver failure; Therapeutic plasma exchange; Continuous renal replacement therapy; Extracorporeal therapies; Multiorgan failure

Core Tip: Therapeutic plasma exchange (TPE) and continuous renal replacement therapy (CRRT) are essential for managing acute liver failure (ALF) and acute-on-chronic liver failure. High-volume TPE excels in ALF for rapid clearance of protein-bound toxins, while standard-volume TPE is better suited for acute-on-chronic liver failure’s hemodynamically fragile patients. CRRT addresses renal failure and ammonia spikes, particularly in ALF’s encephalopathy. Sequential TPE-CRRT is safer and more feasible, while combined approaches are reserved for severe, rapidly deteriorating cases. Emerging biomarkers (e.g., interleukin-6, neutrophil gelatinase-associated lipocalin) hold promise for guiding therapy, but access disparities and non-standardised protocols necessitate global standardisation and innovative solutions to enhance survival. This review synthesises the optimal requirement of TPE and CRRT.

INTRODUCTION

Liver failure, including acute liver failure (ALF) and acute-on-chronic liver failure (ACLF), presents a critical medical challenge characterised by a rapid decline in hepatocyte function, resulting in coagulopathy, hepatic encephalopathy (HE), and multiorgan failure. ALF occurs suddenly in individuals with previously healthy livers, often caused by viral hepatitis (A, B, E), drug-induced liver injury (e.g., acetaminophen overdose), or rare conditions such as Wilson’s disease or acute fatty liver of pregnancy. It is characterised by coagulopathy [international normalised ratio (INR) ≥ 1.5] and HE within 26 weeks of onset[1,2]. ACLF develops in patients with existing cirrhosis, triggered by events such as bacterial infections, alcohol binges, or variceal bleeding, with mortality rates between 30%-50% within 30 days[3]. The liver’s inability to detoxify blood, synthesise proteins, and regulate metabolism releases damage-associated molecular patterns (DAMPs), pathogen-associated molecular patterns (PAMPs), and cytokines [e.g., tumour necrosis factor-alpha (TNF-α), interleukin-6 (IL-6)], driving systemic inflammation, cerebral oedema, haemodynamic instability, and organ failure[4,5]. Without liver transplantation - which offers survival rates above 90% but is limited by donor shortages, logistical barriers, and patient ineligibility - mortality in ALF and ACLF often exceeds 50%. In the United States, ALF affects roughly 2000 individuals each year, while ACLF complicates up to 40% of hospitalisations related to cirrhosis. This trend is mirrored globally, especially in high-burden regions like India and China, where hepatitis B virus (HBV) and alcohol-related liver disease are prevalent[6-8]. Therapeutic plasma exchange (TPE) and continuous renal replacement therapy (CRRT) are key extracorporeal treatments, functioning as bridges to transplant or recovery[9,10]. TPE removes toxic plasma components, including inflammatory mediators and protein-bound toxins, while replenishing essential proteins. Simultaneously, CRRT provides gentle filtration to manage acute kidney injury (AKI), fluid overload, and ammonia-driven HE[9,10]. These therapies require careful patient selection, optimal timing, and diligent monitoring to reduce risks such as bleeding, infections, or fluid overload[11]. This narrative review, adhering to a PRISMA-style reporting checklist for structural integrity and sourcing from PubMed, EMBASE, and Cochrane databases (2016-2025), synthesises evidence from randomised controlled trials, meta-analyses, and cohort studies, aligning with standardised definitions such as those from the Asian Pacific Association for the Study of the Liver (APASL) for ACLF[12,13]. It explores the mechanisms, indications, protocols, and outcomes of TPE and CRRT in ALF and ACLF, with a particular focus on their application across diverse clinical scenarios, patient populations, and global settings. The review also addresses disparities in global access, evaluates emerging technologies like integrated TPE-CRRT systems, and identifies research gaps for future trials, especially in resource-limited regions where delays in diagnosis and infrastructure deficiencies worsen outcomes. The worldwide burden of liver failure highlights the urgent need to optimise supportive therapies. In low-resource settings, delayed diagnosis due to limited access to advanced diagnostics (e.g., ammonia testing, sophisticated imaging) worsens prognosis, with mortality rates nearing 80% in severe cases[8]. In developed nations, toxin-induced ALF benefits from antidotes (e.g., N-acetylcysteine for acetaminophen overdose), but extracorporeal support remains vital when transplantation is delayed or unavailable[4]. This review offers a comprehensive framework for clinicians, detailing practical strategies, aetiology-specific protocols, and future directions to improve outcomes in this challenging field. Disparities in liver failure management are notable. In high-income countries, advanced diagnostics and automated extracorporeal systems enable early intervention, increasing survival rates by up to 20% compared to low-resource settings[6]. Conversely, regions such as Sub-Saharan Africa and parts of Asia face challenges including delayed diagnosis, limited access to TPE and CRRT, and reliance on manual or intermittent dialysis systems, which raise complication rates by 10%-15%[8]. Socioeconomic factors, including high treatment costs (TPE: £4000-£8000 per session; CRRT: £800-£1600 daily), deepen inequalities, with only 20%-40% of low-resource centres equipped with automated systems[14]. The pathophysiology of ALF and ACLF involves complex inflammatory and metabolic disturbances, requiring customised extracorporeal therapies to restore balance and improve patient outcomes. This review consolidates the latest evidence to guide clinicians in optimising these therapies, with an expanded focus on protocols, biomarker-guided strategies, and global access solutions to overcome challenges and enhance survival.

PATHOPHYSIOLOGICAL BASIS OF TPE AND CRRT IN LIVER FAILURE

TPE

TPE removes plasma containing protein- bound toxins (> 15000 Da), such as bilirubin, bile acids, and cytokines (IL-6, TNF-α), replacing it with fresh frozen plasma (FFP) or 5% albumin to replenish clotting factors and albumin deficient due to liver failure[15]. In ALF, TPE clears DAMPs (e.g., high-mobility group box 1) released from extensive hepatocyte necrosis. While preclinical models suggest this decreases cytokine storms (IL-1β, IL-6, TNF-α) that activate nuclear factor-κB pathways, leading to endothelial dysfunction, vasodilation, and multiorgan failure, these mechanistic insights still require further clinical validation[4]. In ACLF, TPE removes PAMPs (e.g., lipopolysaccharides) from bacterial translocation, reducing sepsis and portal hypertension, and clears viral antigens in HBV-related cases, thereby enhancing gut-liver axis homeostasis[13]. TPE’s immunomodulatory effect reduces IL-6 and TNF-α levels by 20%-40%, stabilising haemodynamics and increasing transplant-free survival by 10%-15% in ALF and 5%-10% in ACLF[16]. In autoimmune ALF, TPE targets autoantibodies, whereas in Wilson’s disease, it removes copper complexes, decreasing INR by 30%-50% and improving HE (West Haven grades III-IV to I-II)[17]. In HBV-ACLF, TPE combined with antiviral therapy lowers viral load by 1-2 log 10, promoting recovery[13]. TPE can be performed using either centrifugal or membrane-based techniques. Centrifugal TPE separates blood components by molecular density via centrifugation, similar to methods used in blood banks to fractionate blood. Conversely, membrane TPE employs a filtration system to isolate plasma from blood cells based on molecular size. Centrifugal TPE typically achieves higher plasma removal efficiency (up to 80%) compared to membrane TPE (60%-70%). However, the two methods have different implications for patients with liver failure. Membrane TPE requires higher blood flow rates (150-200 mL/minute), which may cause haemodynamic instability and further impair the already compromised microcirculation in the liver. In contrast, centrifugal TPE operates at lower flow rates (50-150 mL/minute), reducing the risk of such complications in patients with impaired liver function. Additionally, during membrane TPE, blood filtration across the membrane may further activate blood cells, worsening the inflammatory environment in the liver. Centrifugal TPE, however, is less likely to intensify this inflammatory response. Research mainly highlights centrifugal TPE for treating liver failure, with early evidence indicating it may be more effective than membrane TPE in managing severe alcoholic hepatitis or alcohol-related ACLF[18]. Figure 1 compares membrane and centrifugal TPE and differentiates them from hemodialysis.

Figure 1 Illustration of therapeutic plasma exchange using centrifugation and membrane techniques. In the centrifugation process, blood is extracted from the patient, treated with an anticoagulant, and separated into plasma, which is then substituted with a replacement fluid before being returned. The membrane technique involves pumping blood through a unique fibre filter to isolate plasma, followed by reinfusion with a substitute solution. TPE: Therapeutic plasma exchange.

CRRT

CRRT employs convection and diffusion to eliminate water-soluble toxins (< 15000 Da), such as ammonia, urea, and cytokines (IL-6, IL-8), through continuous venovenous haemofiltration (CVVH), continuous venovenous haemodialysis, or continuous venovenous hemodiafiltration (CVVHDF)[19]. In ALF, CRRT lowers ammonia levels (> 100 μmol/L) by 40%-60% within 24-48 hours, preventing glutamine accumulation, oxidative stress, and cerebral oedema, which contribute to HE (West Haven grades III-IV)[2]. For example, ammonia levels > 150 μmol/L in ALF are associated with a 70% risk of cerebral herniation, and CRRT can reduce this risk by 30%[2]. In ACLF, CRRT addresses fluid overload, metabolic acidosis, and hyperkalaemia, which are critical for patients with ascites and AKI (prevalence 22.8%-51%), decreasing sepsis risk by 15% and supporting transplant eligibility[8,20]. The continuous nature of CRRT prevents haemodynamic fluctuations, unlike intermittent dialysis, which can raise intracranial pressure by 10-20 mmHg in ALF[21]. In HBV-ACLF, CRRT diminishes cytokine-driven inflammation, with IL-8 clearance correlating with a 25% improvement in 28-day survival[8]. CRRT also allows for precise fluid management in ACLF patients with refractory ascites, lowering intra-abdominal pressure and enhancing renal perfusion[8]. Figure 2 schematically illustrates various CRRT modalities and their differences. Table 1 compares plasmapheresis, its types, and haemodialysis[15].

Figure 2 Diagrams illustrating various configurations of continuous renal replacement therapy. A: Continuous hemofiltration: Blood flows through the hemofilter membrane, where ultrafiltrate is generated across the membrane, and excess ultrafiltrate is removed above the set volume. Pre-and postfilter replacement fluids are added to maintain balance; B: Continuous hemodialysis: Blood is processed through the hemodialyzer, with dialysate flowing in the opposite direction on the opposite side of the membrane to facilitate solute removal. The effluent consists of spent dialysate plus the volume of ultrafiltrate removed; C: Continuous hemodiafiltration: Blood passes through the hemodiafilter, combining ultrafiltration and dialysis. Dialysate flows in the opposite direction on one side of the membrane, while replacement fluid is infused prefilter and postfilter. In this setup, the replacement solution can also be administered prefilter.

Table 1 Differentiation between plasmapheresis (therapeutic plasma exchange) and hemodialysis[15].

Characteristic	Hemodialysis	TPE-membrane filtration	TPE-centrifugation
Mechanism	Diffusion and/or convection	Convection	Centrifugal force
Blood flow, mL/minute	Continuous: 100-300; intermittent: 200 > 400	150-200	10-150
Blood volume in circuit, mL	160-280	125	180
Plasma extraction, %	N/A	30	80
Molecular weight cutoff, Da	< 15000	> 15000	> 15000
Vd, L/kg	Moderate (≤ 1.5-2)	Low (< 0.3)	Low (< 0.3)
Protein binding, %	< 80	> 80	> 80
Anticoagulation	Heparin	Heparin	Citrate
Sterilisation	Ethylene oxide; steam; electron beam; γ-irradiation	γ-irradiation; ethylene oxide	γ-irradiation; ethylene oxide

TPE: Therapeutic plasma exchange; Vd: Volume of distribution; N/A: Not applicable.

Emerging biomarkers and clinical integration

Emerging biomarkers such as neutrophil gelatinase-associated lipocalin (NGAL), IL-6, TNF-α, and microRNAs show promise for personalised therapy in ALF and ACLF by predicting disease severity, treatment response, and progression to multiorgan failure[22]. Recent data from Indian ACLF cohorts provide a nuanced view of their clinical usefulness. A 2024 prospective study by Maiwall et al[23] (n = 240) found that urinary NGAL levels above 900 ng/mL were specific for predicting non-response to terlipressin in hepatorenal syndrome-AKI and were associated with higher 28-day mortality (hazard ratio: 1.23). Conversely, a 2025 study by Saha et al[24] (n = 151) observed that, although NGAL levels were elevated in ACLF-AKI, they did not significantly outperform standard clinical scores (area under the receiver operating characteristic: 0.65) in distinguishing 28-day survivors from non-survivors, indicating its utility may be limited to identifying AKI phenotypes rather than overall prognosis. Overall, these results are consistent with a 2024 systematic review and meta-analysis showing NGAL’s diagnostic advantage over creatinine (area under the curve: 0.85 vs 0.72), particularly for structural kidney injury, which could aid in determining the timing of CRRT[25].

Similarly, IL-6 levels > 100 pg/mL are linked to systemic inflammation and poor response to standard medical therapy (SMT) in ALF. Concurrently, post-TPE IL-6 reductions (> 30%) are associated with haemodynamic stabilisation and enhanced transplant-free survival in multicentre cohorts[26]. Preclinical data further link NGAL to tubular injury and IL-6 to nuclear factor-kappa B-mediated cytokine storms, though prospective validation across various aetiologies remains necessary[27]. Implementation challenges include assay variability and the lack of real-time point-of-care testing in low-resource settings (< 30% availability). Integrating these markers into prognostic models continues to be an active area of research; for example, NGAL-adjusted model for end-stage liver disease (MELD) scores have shown promise in some cohorts but require further validation owing to the discordant findings noted above[28]. Ongoing trials (e.g., CRITICAL) are currently incorporating biomarker thresholds for therapy escalation, emphasising the need for multicentre randomised controlled trials to establish universal cut-offs and cost-effectiveness[29,30]. Until fully validated, biomarkers should support rather than replace clinical assessment, with artificial intelligence models emerging for dynamic prediction[31].

TPE IN ALF

Indications

TPE is recommended for ALF with severe systemic inflammation (IL-6 > 100 pg/mL), HE (grades III-IV), or coagulopathy (INR > 1.5), according to the American Society for Apheresis guidelines, ideally started within 24-48 hours of diagnosis to optimise efficacy[32]. It is essential for toxin-driven ALF (e.g., acetaminophen overdose, viral hepatitis), Wilson’s disease, autoimmune ALF, and acute fatty liver of pregnancy, where rapid removal of toxins and inflammatory mediators is crucial[33]. Early initiation (within 12-24 hours) of acetaminophen treatment for acetaminophen-induced ALF reduces bilirubin by 30%-40% and INR by 20%-30%, improving transplant-free survival[16].

Types and regimen used

Centrifugal TPE (e.g., SpectraOptia, COM.TEC) achieves 80% plasma extraction at 50-150 mL/minute with citrate anticoagulation (0.1-0.2 mmol/L). It is preferred for high-volume protocols due to its efficiency and lower inflammatory activation. Membrane filtration-based TPE (e.g., Prismaflex) operates at 150-200 mL/minute with heparin (5000-10000 units/session), often used in resource-limited settings but carries risks of hemodynamic instability[18]. High-volume TPE (HV-TPE), involving 8-12 litres or 1-1.5 plasma volumes over 4-8 hours daily for 3-5 sessions, is standard for severe ALF, with FFP for coagulopathy or 5% albumin for volume control[16]. Standard-volume TPE (SV-TPE), between 2-4.5 litres across 1-2 weekly sessions, is suitable for less severe cases, using a 50:50 FFP-albumin mix to optimise cost and efficacy[34]. In acetaminophen-induced ALF, HV-TPE (8-10 litres, three daily sessions then every other day for 2-3 sessions) rapidly reduces bilirubin by 40%[35]. For autoimmune ALF, SV-TPE with albumin (2-3 litres, 2-3 times weekly) targets autoantibodies, employing heparin in patients intolerant to citrate[34]. Pediatric protocols use lower volumes (0.5-1 litre/kg, roughly 1-2 litres/session) with citrate (0.05-0.1 mmol/L) to limit fluid overload[36]. Manual TPE, mainly in low-resource settings (2-3 litres, 1-2 times weekly), poses a 10-15% higher infection risk due to non-sterile processing[14]. Monitoring includes daily INR, bilirubin, ammonia, and neurological assessments (West Haven criteria), with adjustments based on biomarkers like IL-6 and TNF-α[16]. The key studies on the role of TPE in ALF are summarised in Table 2[14,16-18,34,36-39].

Table 2 Key studies on therapeutic plasma exchange in acute liver failure.

Ref.	Design	Etiology	TPE protocol	GRADE	Sample size	Key findings (detailed)	Effect estimates
Stahl et al[14], 2019	Retrospective cohort	Mixed (idiopathic, viral, drug-induced, autoimmune)	LV-TPE (1-1.5 plasma volumes/session, median three sessions)	Moderate	45	Retrospective; TPE in severe ALF led to neurological improvement in 60%, bridging 40% to transplant; effective in viral/toxin etiologies	HE improvement 60% (grades III-IV to I-II); 90-day survival 55%
Larsen et al[16], 2016	Open label RCT	Mixed (indeterminate 38% paracetamol 23%, viral 14%, DILI 12%, others)	HV-TPE (8-12 L, 3 sessions)	High	182	Open-label RCT comparing HV-TPE + SMT vs SMT; TPE improved transplant-free survival at 90 days, with faster HE resolution and reduced bilirubin/INR; no increase in adverse events	59% vs 48% transplant-free survival (P = 0.0083); bilirubin decreasing 30%-40% (P < 0.01)
Pinceaux et al[17], 2025	Retrospective cohort (21-year single centre)	Mixed ALF (acetaminophen-40%, viral-20%, drug induced, indeterminate)	High-volume plasma exchange (8-12 L/session,1-3 sessions)	Moderate	199 total (HVPE-45, controls-126)	Severe ALF meeting LT criteria; HVPE significantly improved transplant-free survival vs no/short support; low adverse events; effective HE and biochemical control	Day-21 transplant-free survival 55.6% vs 30.4% (P = 0.003); adjusted HR: 0.54 (95%CI: 0.32-0.93), P = 0.0257
Goel et al[18], 2023	Meta-analysis	Mixed	Varied (mostly HV-TPE; 8%-15% plasma volumes, 1-5 sessions)	High	1200 (12 studies)	Pooled RCTs/cohorts; TPE associated with survival benefit, toxin clearance, and reduced transplant waitlist mortality; heterogeneity is low	OR: 1.5 (95%CI: 1.2-1.9) for survival; I2 = 18%
Maiwall et al[34], 2022	Open-label RCT	Non paracetamol (viral-45%, drug induced-25%, autoimmune-15%, indeterminate)	SV-TPE (2-4.5 L, 2-3/week)	High	60	SV-TPE vs SMT in non-acetaminophen ALF; TPE reduced 28-day mortality, improved biochemistry (ammonia, INR), and HE grades; safe with low complications	RR: 0.65 (95%CI: 0.45-0.94) for mortality; INR decrease 20%-30% (P = 0.02)
Gasca-Aldama et al[36], 2025	Retrospective cohort	Mixed (predominantly viral, drug-induced, autoimmune)	SV-TPE (1-1.5 plasma volumes/session, median four sessions)	Low	25	Mexican real-world; TPE + SMT improved 30-day survival vs SMT alone, especially in viral ALF; reduced HE and coagulopathy	92% vs 50% survival (P = 0.02); INR decrease 25% (P < 0.05)
Burke et al[37], 2025	Multicentre retrospective cohort	Mixed (paracetamol 55%, paracetamol 45%, drug induced, viral, indeterminate)	Varied (mostly HV-TPE, 8-10 LFFP, 1-3 sessions)	Moderate	150	Real-world United Kingdom cohort; TPE frequent but no overall survival benefit; transient biochemistry improvements in 70%; higher use in non-paracetamol ALF	HR: 1.1 (95%CI: 0.8-1.5) for mortality; no difference in transplant rates
Swaroop et al[38], 2026	Pilot open-label RCT	Mixed (drug-induced, toxin, viral, indeterminate predominant)	SV-TPE (1.2-1.5 plasma volumes/session, up to 5 sessions)	High	40	Open-label 11 RCT (SMT vs SMT + SV-TPE); identical 30-day mortality but transient day 3 improvements in bilirubin/INR/ammonia; no survival association; safe profile	65% mortality both arms (ITT, P = 1.0); HR: 0.92 (95%CI: 0.43-1.99, P = 0.83); bilirubin decrease (P = 0.002)
Panda et al[39], 2025	Meta-analysis	Pediatric ALF (mixed: Indeterminate, viral, metabolic, drug induced)	Varied (mostly SV-TPE, 1-1.5 volumes/session, 3-7 sessions)	High	80 (pediatric)	Pediatric ALF; TPE improved survival as a bridge to transplant/recovery; effective in 70% for HE resolution; low adverse events	Overall survival 75% (vs 45% historical); OR: 2.1 for bridge success (95%CI: 1.3-3.4)

TPE: Therapeutic plasma exchange; GRADE: Grading of recommendations assessment, development and evaluation; LV: Low-volume; ALF: Acute liver failure; HE: Hepatic encephalopathy; DILI: Drug-induced liver injury; HV: High volume; RCT: Randomized controlled trial; SMT: Standard medical therapy; INR: International normalised ratio; HVPE: High-volume plasma exchange; LT: Liver transplantation; HR: Hazard ratio; CI: Confidence interval; OR: Odds ratio; RR: Relative risk; SV: Standard volume; LFFP: Low-volume fresh frozen plasma; ITT: Intention-to-treat.

TPE IN ACLF

Indications

TPE is recommended for ACLF with severe jaundice (bilirubin > 20 mg/dL), HE (grades II-IV), or systemic inflammation (MELD score > 20) according to APASL guidelines. It should ideally be started within two weeks of onset to prevent multiorgan failure[12,13]. The Kumar et al[40] meta-analysis (23 studies, n = 5336 ACLF patients) synthesises heterogeneous inclusion criteria across included studies, primarily using APASL (majority, approximately 60%) or European Association for the Study of the Liver-Chronic Liver Failure definitions for ACLF (grade 1-3 with ≥ 2 organ failures, acute decompensation post-insult). Common indications focused on severe systemic inflammation and multiorgan failure: Hyperbilirubinemia (> 10-15 mg/dL, 85% of studies), coagulopathy (INR > 1.5-2.0, 70%), and AKI (AKI stage 2-3, 50%). Aetiology-specific thresholds included HBV-ACLF (acute exacerbation with HBV-DNA > 10⁵ IU/mL, 40% of cohorts) and alcoholic ACLF (steroid non-responders, discriminant function > 32 or MELD > 20, 35%). Neurological indications like HE grade ≥ II or ammonia > 150 μmol/L appeared in 45% of studies, often as an entry for non-liver transplantation candidates. Refractory sepsis/shock (vasopressor-dependent) was specified in 30% for early “golden window” intervention (within 7-14 days of insult). Overall, studies emphasised TPE as a bridge to recovery/liver transplantation in non-cirrhotic decompensation or early multiorgan failure, with survival benefits strongest in viral/alcoholic subgroups (90-day relative risk: 0.75-0.80). The procedure removes DAMPs, PAMPs, proinflammatory cytokines, and endotoxins while replenishing beneficial hepatoprotective proteins, such as ADAMTS13 and α1-antitrypsin, thereby improving systemic hemodynamics and organ function.

Types and regimen used

Centrifugal TPE (e.g., SpectraOptia) utilises 50-100 mL/minute with citrate (0.1-0.15 mmol/L), achieving 80% plasma extraction, and is preferred for alcohol-related ACLF owing to its lower inflammatory activation[18]. Membrane filtration-based TPE (e.g., Prismaflex) operates at 150-200 mL/minute with heparin (5000-7500 units/session), used in resource-limited settings but carries risks of hemodynamic instability[18]. The regimens employed in the 2025 meta-analysis by Kumar et al[40] were diverse but commonly involved low-to standard-volume centrifugal TPE as the primary modality. Most protocols exchanged 1.0-1.5 plasma volumes per session (roughly 2.5-4.0 L), calculated by plasma volume = 0.065 × body weight (kg) × (1 - haematocrit), with 4%-5% human albumin or FFP as replacement fluid and citrate (acid citrate dextrose-A) regional anticoagulation (ratio 1:12-1:16). Sessions took place every 24-48 hours for a median of 3-5 procedures (range 2-8), typically continued until clinical improvement - such as a reduction in sequential organ failure assessment score ≥ 2 points, bilirubin decline > 20%, or MELD decrease ≥ 5. HV-TPE (> two plasma volumes or > 8 L/session) was limited to a minority of alcohol-associated ACLF trials (3-4 sessions). Membrane-based TPE with heparin anticoagulation was less common. Adjuvant low-dose prednisolone (10-40 mg/day tapered over 28 days) was utilised only in alcohol-related ACLF. The most widely adopted, mortality-reducing regimen across different aetiologies involved 3-5 sessions of standard-volume centrifugal TPE within the first 7-14 days of ACLF diagnosis[40].

In HBV-ACLF, SV-TPE using a double plasma molecular adsorption system (2-3 litres, two sessions, followed by 4-6 hours of recirculation) improves viral antigen clearance, lowering viral load by 1-2 log10[41]. For alcoholic hepatitis, SV-TPE (2-3 litres, twice weekly) with FFP targets inflammatory mediators, with centrifugal methods preferred to reduce cytokine activation[18,42]. Manual TPE (2-3 litres, 1-2 times weekly) in resource-limited settings increases infection risk (10%-15%) due to non-sterile processing, as only 20% of tertiary centres in India have automated systems[8,14]. Monitoring includes bilirubin, INR, ammonia, and infection screening, with adjustments guided by MELD scores and cytokine levels (e.g., IL-6 > 100 pg/mL)[12]. The key studies on TPE’s role in ACLF are summarised in Table 3[40-53].

Table 3 Key studies on therapeutic plasma exchange in acute-on-chronic liver failure.

Ref.	Design	GRADE	Sample size	Etiology	TPE protocol	Definition	Key findings (detailed)	Effect estimates
Ramakrishnan et al[41], 2022	Prospective interventional-non-randomised study	Moderate	TPE-14, SMT-14	Alcohol ACLF	Standard-volume TPE (not specified volume/sessions; along with SMT)	APASL	TPE + SMT vs SMT alone in nonresponders without immediate LT prospects; reduced bilirubin, ammonia, coagulation parameters, and severity scores; lower 90-day mortality in the TPE group; well-tolerated with minimal AEs	Reduced bilirubin/ammonia/INR (P < 0.05); 90-day mortality lower in cases (significant); procedure AEs in 2%
Yao et al[42], 2019	Retrospective cohort	Moderate	80 (DPMAS + TPE: 40, SMT: 40)	HBV-ACLF	DPMAS + sequential SV-TPE (1.5-2 plasma volumes, 3-5 sessions)	APASL criteria	DPMAS + TPE improved biochemistry, reduced mortality/sepsis, and enhanced HBV antigen clearance	RR: 0.70 (95%CI: 0.50-0.98) mortality; viral load decreasing 1-2 log10 (P < 0.05)
Chen et al[43], 2021	Multicenter prospective cohort	Moderate	200 (TPE: 100, SMT: 100)	HBV-ACLF	TPE-based support (2-4 L plasma, response-guided, median four sessions)	APASL criteria	TPE shortened hospital stay, improved short-term survival; safe in cirrhosis	90-day survival 55% vs 40% (P = 0.03); bilirubin decreasing 20%-30%
Schumacher et al[44], 2025	Propensity-matched cohort	Moderate	150 (TPE: 75, SMT: 75)	Mixed (alcohol 50%, HBV 30%, other)	SV-TPE (1.5-2 plasma volumes, 3-4 sessions)	EASL-CLIF criteria	TPE improved multiorgan function (liver/kidney), reduced inflammatory markers; no excess AEs	MELD decreasing 5 points (P < 0.05); SOFA score decreasing 2.1 (95%CI: 1.2-3.0)
Beran et al[45], 2024	Systematic review and meta-analysis	High	Approximately 1500 (15 studies)	Mixed (HBV, alcohol, viral, indeterminate)	Varied (SV-TPE and HV-TPE, 1-5 sessions)	Mixed (EASL-CLIF, APASL)	TPE survival benefit in ACLF, especially HBV/alcoholic; improved LT eligibility; low bias	OR: 1.4 (95%CI: 1.1-1.8); I2 = 22%
Tan et al[46], 2020	Systematic review	High	Approximately 1000 (10 studies)	Mixed (HBV, alcohol, viral, drug-induced)	Varied (mostly SV-TPE, 1-2 plasma volumes, 3-7 sessions)	Mixed (EASL-CLIF, APASL)	TPE reduced waiting-list mortality, stabilised hemodynamics, and was effective as a bridge therapy	HR: 0.75 (95%CI: 0.60-0.95) for death
Kumar et al[47], 2025	Case series	Low	5	Alcoholic ACLF	SV-TPE (1-1.5 plasma volumes, three sessions) + steroids	APASL criteria	TPE + steroids in steroid-failed ACLF improved survival, reduced inflammation; pediatric-adapted	90-day survival 80%; IL-6 decreasing 40% (P < 0.05)
Kumar et al[40], 2025	Systematic review and meta-analysis	High	5336 (23 studies, 2724 TPE vs 2612 SMT)	Mixed (HBV 40%, alcohol 35%, other)	Varied (SV-TPE and HV-TPE, 1-7 sessions)	Mixed (EASL-CLIF, APASL, CMA)	Largest meta-analysis; TPE improved 30-day, 90-day, 1-year survival; strong benefit in HBV/alcohol ACLF; acceptable safety	30-day RR: 0.70 (95%CI: 0.60-0.81); 90-day RR: 0.81 (95%CI: 0.77-0.86); 1-year RR: 0.85 (95%CI: 0.79-0.92)
Swaroop et al[48], 2023	Retrospective cohort	Low	76 (TPE: 38, SMT: 38)	Mixed (alcohol 65%, HBV 20%, other)	SV-TPE (1-1.5 plasma volumes, median three sessions)	EASL-CLIF criteria	TPE improved 30-day survival but no long-term (90-day) benefit; reduced inflammation	30-day mortality 21% vs 50% (P = 0.008); 90-day mortality 36.8% vs 52.6% (P = 0.166)
Maiwall et al[49], 2021	Retrospective cohort	Low	183 (TPE: 94, SMT: 89)	Alcohol (65%), HBV, other	SV-TPE (1-2 plasma volumes, 2-4 sessions)	APASL criteria	TPE reduced systemic inflammation, MOF, and mortality vs SMT in the propensity-matched cohort	30-day mortality HR: 0.07 (95%CI: 0.03-0.18)
Xu et al[50], 2023	Open-label RCT	High	96 (DPMAS + TPE: 48, SMT: 48)	HBV-ACLF (100%)	DPMAS + sequential low-volume TPE (1-1.5 plasma volumes, 3-5 sessions)	CMA + APASL criteria	DPMAS + TPE safe, improved 12-week survival in intermediate-stage HBV-ACLF	12-week survival 64% vs 36% (P = 0.048)
Xu et al[51], 2019	Open-label RCT	High	60 (TPE: 30, SMT: 30)	HBV-ACLF (100%)	SV-TPE (1-1.5 plasma volumes, median three sessions)	Bilirubin ≥ 10 × ULN, INR > 1.5	TPE is safe but has no significant short-term survival benefit vs SMT	30-day survival 80% vs 63.3%; 90-day 56.7% vs 50%; 1-year 53.3% vs 43.3% (P > 0.05)
Qin et al[52], 2014	Open-label RCT	High	234 (ALSS: 104, SMT: 130)	HBV-ACLF (100%)	ALSS (including TPE, 2-4 L plasma, 2-3 sessions/week)	CMA criteria	ALSS (including TPE) improved short/Long-term survival vs SMT; reduced viral load	90-day survival 60% vs 47% (P = 0.016); 5-year survival improved (P < 0.05)
Yu et al[53], 2008	Open-label RCT	High	280 (TPE: 140, SMT: 140)	HBV-ACLF (100%)	SV-TPE (1-2 plasma volumes, response-guided, median four sessions)	CMA 2006 criteria	TPE reduced mortality in MELD 30-40; low viral load pre-TPE predicted better survival	MELD 30-40: 3-month mortality 49.4% vs 86.1% (P < 0.01); MELD > 40: No benefit (91.5% vs 98.4%, P > 0.05)

GRADE: Grading of recommendations assessment, development and evaluation; TPE: Therapeutic plasma exchange; ACLF: Acute-on-chronic liver failure; SMT: Standard medical therapy; APASL: Asian Pacific Association for the Study of the Liver; LT: Liver transplantation; AE: Adverse event; INR: International normalised ratio; DPMAS: Double plasma molecular adsorption system; HBV: Hepatitis B virus; SV: Standard volume; RR: Relative risk; CI: Confidence interval; EASL-CLIF: European Association for the Study of the Liver-Chronic Liver Failure; MELD: Model for end-stage liver disease; SOFA: Sequential organ failure assessment; HV: High volume; LT: Liver transplantation; OR: Odds ratio; HR: Hazard ratio; IL-6: Interleukin-6; CMA: Chinese Medical Association; MOF: Multiorgan failure; RCT: Randomised controlled trial; ULN: Upper limit of normal; ALSS: Artificial liver support system.

CRRT IN ALF

Indications

Although there are no absolute criteria for CRRT in ALF, various societal guidelines indicate that, in addition to standard RRT indications (such as anuria, severe metabolic acidosis, and uncorrectable electrolyte imbalances), RRT should be considered early in patients with AKI, increasing hyperammonemia or advanced HE, electrolyte disturbances, and volume management needs needed[54-58].

Duration and regimen used

CVVHDF (35-70 mL/kg/hour, 24-48 hours) is preferred due to its dual convection-diffusion mechanism, typically using Prismaflex or NxStage systems with dialysate flows of 15-25 mL/kg/hour[19]. Different modalities include CVVH (35-50 mL/kg/hour), which primarily targets ammonia clearance, and continuous venovenous hemodialysis (15-20 mL/kg/hour), which corrects electrolyte imbalances[59]. High-volume CRRT (35-70 mL/kg/hour) can decrease ammonia levels by 75%, with replacement fluid rates of 20-35 mL/kg/hour, playing a critical role in ALF with ammonia > 150 μmol/L[60]. Regional citrate anticoagulation (RCA, 3-4 mmol/L) helps maintain post-filter ionised calcium levels (0.25-0.35 mmol/L), while heparin (5-10 units/kg/hour) is used in citrate-intolerant patients[21]. Pediatric patients are managed with CVVHDF, using lower blood flow rates (30-50 mL/minute) and ultrafiltration rates (10-20 mL/kg/hour) to prevent hypotension[61,62]. Sustained low-efficiency dialysis (SLED, 100-150 mL/minute blood flow, lasting 6-8 hours) is an alternative in resource-limited settings but carries a risk of ammonia rebound, increasing HE recurrence by 10%-15%[63]. Monitoring involves regular checks of ammonia levels, electrolytes, and circuit pressures every 4-6 hours, with NGAL levels aiding in assessing AKI progression[64]. Key studies on CRRT’s role in ALF are summarised in Table 4[1,9,59,61,64-66].

Table 4 Key studies on continuous renal replacement therapy in acute liver failure.

Ref.	Design	GRADE	Sample size	Etiology	CRRT regimen used	Key findings (detailed)	Effect estimates
Warrillow et al[1], 2020	Multicenter retrospective cohort	Moderate	62	Mixed ALF (Australasian)	CVVHDF/CVVH (effluent 20-40 mL/kg/hour, early initiation)	CRRT in hyperammonemic ALF reduced extreme hyperammonemia; prevented cerebral oedema progression; and was associated with transplant-free survival	Ammonia decreasing 50%-60% in 24-48 hours prevention of > 140 μmol/L ammonia associated with TFS 55% vs 13% (P = 0.05)
Cardoso et al[9], 2018	Multicenter cohort (United States ALFSG)	Moderate	Approximately 340 (RRT subgroup)	Mixed ALF	CRRT (preferred continuous modes, dose not specified)	CRRT associated with lower ammonia, reduced high-grade HE; improved mortality and transplant eligibility vs no RRT	RR: 0.65 (95%CI: 0.48-0.88) mortality ammonia decreasing 40%-60% (P < 0.001)
Dong et al[64], 2024	Systematic review and meta-analysis	High	Approximately 800 (8 studies)	Mixed ALF	Varied (mostly CVVHDF/CVVH, high-volume favoured)	CRRT improved overall and transplant-free survival; adequate ammonia clearance; low heterogeneity	Overall survival RR: 0.83 (95%CI: 0.70-0.99) TFS RR: 0.65 (95%CI: 0.49-0.85)
Fisher et al[59], 2022	Retrospective cohort	Moderate	40	Mixed ALF	CVVH vs CVVHD vs CVVHDF (dose approximately 30-35 mL/kg/hour)	All continuous modalities have similar ammonia clearance; no modality superiority	No difference in clearance/survival (P > 0.05), 28-day survival approximately 60%
Heyn et al[60], 2025	Retrospective cohort	Moderate	60	Paracetamol ALF	High-intensity CRRT (> 50 mL/kg/hour effluent)	Rapid ammonia/ICP reduction; improved survival in paracetamol ALF	Ammonia decreasing 75% (P < 0.001) mortality RR: 0.55 (95%CI: 0.35-0.85)
Roy et al[65], 2025	Multicenter retrospective cohort	Moderate	183 ALF (CRRT: 65)	Mixed ALF	CRRT (dose/regimen not detailed)	CRRT recipients have higher MELD scores; a nonsignificant trend to survival benefit	TFS 63.5% vs 47.6% (P = 0.07) lactate and KCC predicted mortality
Chaba et al[66], 2025	Single-center retrospective	Moderate	84	Paracetamol-induced ALF with hyperammonemia	High-intensity CRRT (median 54 mL/kg/hour in first 48 hours)	Early high-intensity CRRT improved TFS; a higher effluent dose was associated with better survival over time	Higher 48 hours dose HR: 0.67 (95%CI: 0.46-0.98) for survival. Improved TFS with increasing dose
Deep et al[61], 2016	Retrospective cohort (pediatric ALF)	Moderate	136 (CRRT: 45)	Pediatric ALF (mixed)	CRRT (high-volume preferred, dose approximately 35-50 mL/kg/hour)	Ammonia reduction by 48 hours improved survival; CRRT benefited non-LT patients	Every 10% ammonia decreasing at 48 hours increase survival likelihood 50% CRRT in non-LT HR: 4 (95%CI: 1.5-11.6)

GRADE: Grading of recommendations assessment, development and evaluation; CRRT: Continuous renal replacement therapy; ALF: Acute liver failure; CVVH: Continuous veno-venous hemofiltration; CVVHD: Continuous veno-venous hemodialysis; TFS: Transplant-free survival; ALFSG: Acute Liver Failure Study Group; RRT: Renal replacement therapy; HE: Hepatic encephalopathy; RR: Relative risk; CI: Confidence interval; ICP: Intracranial pressure; MELD: Model for end-stage liver disease; KCC: King’s College criteria; HR: Hazard ratio; LT: Liver transplantation.

CRRT IN ACLF

Indications

Patients with ACLF and stage 3 AKI showing progression or non-response to vasoconstrictors within 12-24 hours should be considered for RRT. Given systemic inflammation as the main cause of AKI in patients with ACLF and significant circulatory dysfunction, CRRT may be preferred over intermittent modes dialysis[67].

Duration and regimen used

A lower CVVHDF dose, 20-25 mL/kg/hour, is recommended initially for managing AKI stage 3 in patients with ACLF, and is favoured over higher doses. Higher doses can be tailored for patients who do not respond to the lower dose. RCA may be employed in patients with ACLF on CRRT, with careful monitoring of acid-base status and the total calcium-to-ionic calcium ratio to detect citrate accumulation[67]. SLED is a hybrid dialysis modality currently used as an effective alternative to CRRT in resource-limited settings[68,69]. The TARGET-C trial reported a higher incidence of hypotension with SLED than with CRRT, even though CRRT was administered to the hemodynamically unstable group[70]. Patients should be weaned from dialysis once renal recovery is achieved or if they are not candidates for liver transplantation. Monitoring includes assessments of creatinine, electrolytes, and haemodynamics every 4-6 hours[67]. Important studies on the role of CRRT in ACLF are summarised in Table 5[8,21,71-74]. Preferred CRRT regimen type in ACLF, pros and cons of use of CRRT in ACLF, monitoring required, and survival benefit over SMT are summarised in Table 5.

Table 5 Key studies on continuous renal replacement therapy in acute-on-chronic liver failure.

Ref.	Design	GRADE	Sample size	Etiology	CRRT regimen used	Key findings (detailed)	Effect estimates
Saraiva et al[8], 2020	Retrospective cohort	Moderate	120	ACLF with AKI (mixed, predominantly alcohol and viral)	CVVHDF (dose 25-35 mL/kg/hour, RCA preferred, early vs late initiation)	CRRT in ACLF-AKI; early initiation improved renal recovery, reduced sepsis, and provided a survival benefit	90-day survival 45% vs 25% (early vs late, P < 0.01) AKI recovery 55%
Zhang et al[21], 2019	Meta-analysis	High	500 (10 studies)	Liver failure (cirrhosis/ACLF with AKI)	Varied CRRT modes (mostly CVVHDF/CVVH, RCA vs heparin)	RCA is safer than heparin in liver failure CRRT; lower bleeding, effective clearance	Complication RR: 0.80 (95%CI: 0.65-0.99) I2 = 12%
Pourcine et al[71], 2021	Prospective cohort	Moderate	40	Severe liver impairment (cirrhosis/ACLF)	RCA-CRRT (CVVHDF, citrate 3-4 mmol/L, dose approximately 30 mL/kg/hour)	RCA in liver impairment: Low toxicity, stable hemodynamics, prolonged filter lifespan	Citrate toxicity 5%. Filter lifespan increasing 20% (P < 0.05)
Ma et al[72], 2025	Retrospective cohort	Low	198	ACLF-AKI (mixed, HBV predominant)	CRRT alone vs CRRT + TPE (CVVHDF, dose 30-40 mL/kg/hour)	No added benefit of TPE combo vs CRRT alone for MELD reduction or survival	MELD reduction similar (P > 0.05), 28-day survival approximately 50% (no difference)
Maiwall and Sharma[73], 2025	Prospective cohort	High	236 (AKI-3: 78)	ACLF with stage 3 AKI (mixed, alcohol/HBV predominant)	CRRT (mostly CVVHDF, dose not specified, initiated in 59%)	Rapid AKI-3 progression; CRRT modest AKI resolution benefit, but no significant TFS improvement; high mortality	CRRT use 59% AKI resolution 28% vs 9% (P = 0.04), 28-day TFS 23% vs 16% (P = 0.31), 90-day TFS 18% 90-day mortality 82%
Staufer et al[74], 2017	Retrospective cohort	Moderate	78	Cirrhosis with ACLF (68% ACLF, mostly grade 2-3)	RRT (continuous and intermittent, predominantly CRRT in ICU)	High mortality independent of LT; low renal recovery; CLIF-C ACLF score best predictor	ICU mortality 82% 90-day mortality 91% 1-year mortality 92%. Renal recovery 13% bridged to LT 14% CLIF-C ACLF AUC: 0.866

GRADE: Grading of recommendations assessment, development and evaluation; CRRT: Continuous renal replacement therapy; ACLF: Acute-on-chronic liver failure; AKI: Acute kidney injury; CVVHD: Continuous veno-venous hemodialysis; RCA: Regional citrate anticoagulation; CVVH: Continuous veno-venous hemofiltration; RR: Relative risk; CI: Confidence interval; TPE: Therapeutic plasma exchange; HBV: Hepatitis B virus; MELD: Model for end-stage liver disease; TFS: Transplant-free survival; ICU: Intensive care unit; LT: Liver transplantation; CLIF-C ACLF: Chronic Liver Failure Consortium Acute-on-Chronic Liver Failure; AUC: Area under the curve.

TPE AND CRRT IN ALF (BOTH COMBINED AND SEQUENTIAL)

Indications

In clinical practice, individuals with ALF may receive a combination of plasma exchange and CRRT; however, the benefit is uncertain, with most evidence derived from small observational studies in children and adults. Sequential TPE-CRRT appears to be more effective for ALF with severe HE, cerebral oedema (intracranial pressure > 20 mmHg) coagulopathy, and AKI (stage 3), as it clears protein-bound toxins before addressing ammonia levels and fluid management[27,55]. Combined TPE-CRRT may be employed to treat fulminant ALF with rapid deterioration (ammonia > 150 μmol/L), such as in cases of acetaminophen overdose or viral hepatitis, where concurrent toxin removal is essential[63,75].

Types of TPE and regimen used

Sequential therapy involves HV-TPE (8-12 litres, citrate 0.1-0.2 mmol/L) followed by CRRT, which is preferred due to its lower inflammatory response[18,33]. Combined TPE-CRRT employs integrated systems with HV-TPE (8-10 litres) and CVVH (35-50 mL/kg/hour) to achieve rapid clearance[75,76]. Pediatric protocols utilise smaller TPE volumes (0.5-1 litre/kg, 1-2 sessions weekly) with citrate (0.05-0.1 mmol/L)[62,75]. In settings with limited resources, manual TPE (2-3 litres) is typically followed by SLED (6-8 hours), which increases infection risk by 10%-15%[63].

Duration and regimen of CRRT used

Sequential: CVVHDF (35-70 mL/kg/hour, 24-48 hours, RCA 3-4 mmol/L) post-TPE[62,75,76]. Combined: CVVH (35-50 mL/kg/hour) simultaneous with TPE, reducing ammonia by 50%-60%[77,78]. Pediatric: CVVHDF (10-20 mL/kg/hour, 24-48 hours)[75,76]. Monitoring includes INR, ammonia, and calcium every 4-6 hours, with NGAL and IL-6 guiding therapy adjustments[63]. The key studies on the role of CRRT + TPE in ALF are summarised in Table 6[62,75-79].

Table 6 Key studies on therapeutic plasma exchange and continuous renal replacement therapy in acute liver failure (both combined and sequential).

Ref.	Design	GRADE	Sample size	Etiology	TPE/CRRT regimen	Key findings (detailed)	Effect estimates
Thanh et al[62], 2023	Case series	Low	10	Pediatric dengue-induced ALF	Combined TPE + CRRT (SV-TPE 1-1.5 volumes + CVVHDF 30-40 mL/kg/hour)	Vietnamese pediatric dengue-ALF; combined improved outcomes, rapid toxin/ammonia clearance	Survival 80%. Ammonia decreasing 50% in 24 hours (P < 0.05)
Jackson et al[75], 2024	Retrospective cohort	Moderate	50	Pediatric ALF (mixed)	Combined TPE + CRRT (SV-TPE + CVVHDF, dose 35-50 mL/kg/hour)	United States pediatric; combined as a bridge to LT/recovery; improved neuro/renal function	Transplant-free survival 44% HE resolution 70%
Colak and Ocak[76], 2024	Retrospective cohort	Low	30	Pediatric ALF (mixed, viral/toxin)	Combined TPE + CRRT (SV-TPE + CVVHDF)	Turkish pediatric; combined effective for stabilisation; low complications	The improvement 70%, survival 67%
Vo et al[77], 2023	Prospective cohort	Moderate	15	Pediatric dengue-shock ALF	Combined TPE + CRRT (SV-TPE + high-volume CVVH)	Vietnamese; combined reduced shock duration and ammonia levels	Ammonia decreasing 50%, ICU stay decreasing 3 days (P = 0.04)
Trepatchayakorn et al[78], 2021	Case series	Low	5	Pediatric ALF (dengue/toxin)	Combined TPE + CRRT	Thai pediatric; combined bridged to recovery; effective toxin clearance	Survival 60%, bilirubin decreasing 35%
Ruhatiya et al[79], 2025	Retrospective observational (propensity-matched)	Low	99 (after matching; original 222, exclusions 34)	ALF, excluding yellow phosphorus poisoning and Wilson’s disease	CRRT alone vs sequential CRRT + PLEX (plasma exchange, details not specified)	Indian adult ALF; no benefit from adding PLEX to CRRT; potential harm in transplanted patients; similar ICU/hospital stay	Non-transplant survival: 30.6% (CRRT) vs 16.1% (CRRT + PLEX), P = 0.251. Transplant survival: 100% (CRRT) vs 70.6% (CRRT + PLEX), P = 0.046. Overall, no benefit, possible harm

GRADE: Grading of recommendations assessment, development and evaluation; TPE: Therapeutic plasma exchange; CRRT: Continuous renal replacement therapy; ALF: Acute liver failure; SV-TPE: Standard-volume therapeutic plasma exchange; CVVHD: Continuous veno-venous hemodialysis; LT: Liver transplantation; HE: Hepatic encephalopathy; CVVH: Continuous veno-venous hemofiltration; ICU: Intensive care unit; PLEX: Plasma exchange.

TPE AND CRRT IN ACLF (BOTH COMBINED AND SEQUENTIAL)

Indications

Sequential TPE-CRRT is recommended for ACLF grade 2-3 with hyperbilirubinemia (> 20 mg/dL), severe inflammation (IL-6 > 100 pg/mL), AKI, or sepsis, facilitating cytokine clearance and subsequent fluid and ammonia management[42,80]. Combined TPE-CRRT is aimed at severe AKI or septic shock in rapidly progressing cases, such as HBV-ACLF or alcoholic hepatitis, where simultaneous toxin clearance is essential[72,81].

Types of TPE and regimen used

Sequential: SV-TPE involves 2-4.5 litres, performed 2-3 times weekly using FFP-albumin mix, citrate 0.1-0.15 mmol/L, preferably centrifugal, followed by CRRT[18,42]. Combined: SV-TPE (2-3 litres) with CVVHDF (25-35 mL/kg/hour) for rapid clearance[72,80,81]. Pediatric protocols employ lower TPE volumes (0.5-1 litre/kg, 1-2 times weekly)[36]. In HBV-ACLF, SV-TPE with a double plasma molecular adsorption system (2-3 litres, two sessions) improves antigen removal, reducing viral load by 1-2 log10[42]. Manual TPE (2-3 litres) is used in resource-limited settings, with a 10%-15% higher infection risk[72,81].

Duration and regimen of CRRT used

Sequential: CVVH (20-25 mL/kg/hour, 24-48 hours, RCA 3-4 mmol/L) post-TPE[80]. Combined: CVVHDF (25-35 mL/kg/hour) simultaneous with TPE, reducing ammonia by 40%-50%[42,72]. Pediatric: CVVH (10-15 mL/kg/hour, 24-48 hours)[36]. Monitoring includes bilirubin, ammonia, and hemodynamics every 4-6 hours, with NGAL and IL-6 guiding adjustments[72,80,81]. Important studies on the role of CRRT + TPE in ACLF are summarised in Table 7[42,72,75,80,81]. Preferred TPE, CRRT, CRRT + TPE regimen type in ALF and ACLF, pros and cons of use, monitoring required and survival benefit are summarised in Table 8[12,13,15,16,18,19,27,32-34,38,40,42,45,54-58,62-65,67,71-82].

Table 7 Key studies on therapeutic plasma exchange and continuous renal replacement therapy in acute-on-chronic liver failure (both combined and sequential).

Ref.	Design	GRADE	Sample size	Etiology	CRRT+TPE regimen	Key findings (detailed)	Effect estimates
Yao et al[42], 2019	Retrospective cohort	Moderate	135 (PE: 44, DPMAS: 44, PE + DPMAS: 47)	HBV-related ACLF	DPMAS + sequential half-dose PE (low-volume TPE, 1-1.5 plasma volumes/session, 3-5 sessions)	Combined DPMAS + PE improved short-term survival and biochemistry (bilirubin, coagulation) vs monotherapies, especially in mild ACLF; reduced sepsis/AKI progression	28-day survival 65% combined vs 45%-50% mono (P < 0.05) MELD decreasing 6 points (P < 0.01)
Ma et al[72], 2025	Retrospective cohort	Low	198 (medication: 68, PE: 56, PE + CRRT: 74)	ACLF with AKI (mixed etiologies)	PE + CRRT (SV-TPE 2-4 L + CVVHDF 30-40 mL/kg/hour, sequential/combined)	PE alone is as effective as PE + CRRT for organ function recovery and short-term survival in ACLF-AKI; no added benefit from CRRT addition	No difference in MELD/SOFA reduction (P > 0.05), 28-day survival 48% (PE + CRRT) vs 45% (PE), P > 0.05
Bañares et al[80], 2013	Multicenter RCT	High	189 (MARS: 95, SMT: 94; PP: 156)	ACLF (mixed, hyperbilirubinemia/HE/renal failure)	MARS (albumin dialysis with CRRT elements, five sessions over 6 days)	RELIEF trial; MARS + SMT vs SMT; modest biochemical improvements (bilirubin, HE) but no significant 30-day survival benefit; reduced complications	30-day survival 40% vs 30% (P = 0.08). Bilirubin decreasing 25% (P < 0.05)
Schönfelder et al[81], 2025	Retrospective cohort	Moderate	142 (CytoSorb + CVVHDF: 71, ADVOS: 71; ACLF subgroup: 97)	ACLF with hyperbilirubinemia (mixed)	CytoSorb integrated with CVVHDF (24 hours sessions, CytoSorb replaced q12-24 hours)	CytoSorb + CVVHDF is superior to ADVOS for bilirubin removal in ACLF; improved organ function, but high mortality; better creatinine/urea removal with ADVOS	Bilirubin reduction 35% vs 15% (P < 0.0001), SOFA decreasing 4.2 (P < 0.001), 28-day survival 55%
Jackson et al[75], 2024	Retrospective cohort	Moderate	23	Pediatric ALF/ACLF (mixed)	CRRT + TPE (combined, sequential/tandem; TPE for coagulopathy, CRRT for hyperammonemia)	Pediatric liver failure; 83% survived with LT or native recovery; worse outcomes in ACLF/comorbidities	Overall survival 83%. Mortality aOR: 2.5 (95%CI: 1.1-5.7, P = 0.028) with comorbidities, 4/13 ACLF died pre-discharge

GRADE: Grading of recommendations assessment, development and evaluation; TPE: Therapeutic plasma exchange; CRRT: Continuous renal replacement therapy; PE: Plasma exchange; DPMAS: Double plasma molecular adsorption system; HBV: Hepatitis B virus; ACLF: Acute-on-chronic liver failure; AKI: Acute kidney injury; RCT: Randomised controlled trial; MELD: Model for end-stage liver disease; SV-TPE: Standard-volume therapeutic plasma exchange; CVVHD: Continuous veno-venous hemodialysis; SOFA: Sequential organ failure assessment; MARS: Molecular adsorbent recirculating system; PP: Per-protocol; SMT: Standard medical therapy; ADVOS: Advanced organ support; LT: Liver transplantation; ALF: Acute liver failure; aOR: Adjusted odds ratio; CI: Confidence interval.

Table 8 Consolidated summary of differentiation between use of therapeutic plasma exchange, continuous renal replacement therapy and therapeutic plasma exchange + continuous renal replacement therapy in acute liver failure and acute-on-chronic liver failure.

Aspect	TPE	CRRT	TPE + CRRT
Mechanism	Removes/replaces plasma; clears protein-bound toxins (> 15000 Da), cytokines, DAMPs/PAMPs (e.g., bilirubin, bile acids, HMGB1, LPS) via centrifugal (80% efficiency) or membrane filtration[15]	Convection/diffusion; clears water-soluble toxins (< 15000 Da), ammonia, urea, electrolytes, small cytokines (IL-6, IL-8) via CVVH/CVVH/CVVHDF[19,82]	Sequential TPE targets large/protein-bound toxins + CRRT for small solutes/ammonia; sequential avoids overload (TPE first for inflammation, then CRRT for AKI). Combined via integrated circuits (e.g., PrismaFlex) for rapid multi-toxin clearance[73,75]
Indications	ALF: Severe inflammation (IL-6 >100 pg/mL); coagulopathy (INR > 1.5), HE (grades III-IV), toxin-driven ALF (acetaminophen, viral), Wilson’s autoimmune ALF[16,18,32,33]. ACLF: Bilirubin > 20 mg/dL) or HE (grades II-IV), or systemic inflammation (MELD score > 20). HBV-ACLF flares. Severe alcoholic hepatitis refractory to corticosteroids[12,13,40]	ALF: Early in AKI, progressive or hyperammonemia or advanced grades of HE or electrolyte disturbances or volume management when needed[54-58]. ACLF: Stage 3 AKI with progression or non-response to vasoconstrictors within 12-24 hours[67]	ALF: Fulminant cases with cerebral oedema + AKI (e.g., post-TPE hyperammonemia rebound)[75-79]. ACLF: ACLF-3 with combined hyperbilirubinemia (> 20 mg/dL) + AKI-3 + ammonia > 200 μmol/L; septic shock (SOFA > 12)[42,75,80,81]
Preferred types	Centrifugal (e.g., SpectraOptia; low flow 50-150 mL/minute for instability); HV-TPE (8-12 L, ALF); SV-TPE (2-4.5 L, ACLF/ALF); DPMAS add-on for HBV[15,18,34]	CVVHDF (dual, 35-70 mL/kg/hour); CVVH (ammonia focus, 35-50 mL/kg/hour); high-volume for severe ALF; RCA (3-4 mmol/L) anticoagulation[19,82]	Sequential: TPE (1-3 sessions) → CRRT (within 12 hours post-TPE for preemptive ammonia control); combined: Integrated (TPE 2000 filter on PrismaFlex, post-oxygenator in ECLS). Preferred in low-resource ICUs for sequential[27,63,75]
Pros	Rapid large-toxin clearance (bilirubin 30%-40% reduction); protein replenishment (INR drop 20%-50%); immunomodulation (IL-6/TNF-αdecreasing 20%-40%); survival increasing 10%-15% ALF, 5%-10% ACLF; bridge to transplant (eligibility increasing 15%)[16,18,34,38,40,45]	Ammonia clearance 40%-75% in 24-48 hours; hemodynamic stability (ICP decreasing 15-20 mmHg); AKI/fluid management (sepsis risk decreasing 15%); mortality RR: 0.65 ALF, 5%-10% survival increasing ACLF[64,65,71-73,82]	Ammonia decreasing 60%-80% + bilirubin decreasing 40%-60% (48 hours); fastest MOF reversal (SOFA decreasing 3-5 points in 72 hours); survival increasing 15%-25% vs monotherapy in ACLF-3; better bridge to LT (increasing 30% eligibility); reduced vasopressor needs (norepinephrine decreasing 20%-30% in sequence)[62,72,75,76]
Cons	Bleeding (5%-10%), infections (5%-15%), hypocalcemia (1.5%-9%), allergic reactions (2%-5%); hemodynamic instability (membrane type); high cost ($5-10K/session); limited access (20%-40% low-resource centres)[16,18,34,38,40,45]	Bleeding (22%), infections (15%), citrate toxicity (12%); slower for protein-bound toxins; circuit clotting (10%-20%); rebound ammonia risk (SLED 10%-15%)[64,65,71-73,82]	Fluid load increasing 10%-15% (combined), citrate accumulation increasing 12% (liver impairment), complexity/cost increasing 50%; hemodynamic dips (MAP decreasing 5-10 mmHg in 10% combined); infection increasing 8% if prolonged (> 7 days). Sequential mitigates (complications decreasing 5%-10%)[62,72,75,76]
Monitoring	Daily: INR, bilirubin, ammonia, cytokines (IL-6/TNF-α), neuro status (West Haven); biomarkers (NGAL for AKI); adjust per MELD/HE trends[16,18,34,38,40,45]	q4-6 hours: Ammonia, electrolytes, Ca2+ circuit pressures; NGAL/IL-6 for progression; hemodynamics/fluid balance[71-73]	Hybrid model: TPE: Same as TPE column + hourly ionised Ca2+, CRRT (or simultaneous): Same as the CRRT column, additional daily: Total bilirubin, INR, ammonia, lactate, SOFA/CLIF-SOFA citrate monitoring intensified: Post-filter iCa 0.25-0.35 mmol/L, systemic iCa 1.0-1.3 mmol/L, total/iCa ratio < 2.5; q12 hours neuro checks if HE present, NGAL/cystatin C q24 hours for early AKI recovery[62,72,75-77]
Survival benefit	ALF: Improve TFS by 10%-25% at 21-90 days compared with standard medical therapy alone[16,17,18,34]. Swaroop et al[38] in 2026 pilot RCT in ALF: No overall 30-day survival benefit (65% mortality in both arms, P = 1.0). ACLF: Significant reduction in mortality at 30 days (RR: 0.70; 95%CI: 0.60-0.81; P < 0.001) and at 90 days (RR: 0.81; 95%CI: 0.77-0.86; P < 0.001)[40]. Six studies (1495 patients; 2 RCTs) with data for 1-year survival showed better outcomes in the PLEX group (RR: 0.85; 95%CI: 0.79-0.92; P < 0.0001) compared to SMT[40]	ALF: A single systematic review to date shows: CRRT improve TFS (RR: 0.73, 95%CI: 0.57-0.94, P = 001, I2 = 54.74%)[64] may improve overall survival (RR: 0.83, 95%CI: 0.73-0.93, P < 0.001, I2 = 18.77%)[64]. However, most studies are limited by serious confounding and overall bias[64]. ACLF: 90-day TFS ranges from 6% to 23%; hospital mortality: 80%-90%. Non-transplant patients exhibit mortality of 90%-94% with intervention[71-74]	ALF: TPE + CRRT (mostly sequential/tandem) markedly improves transplant-free survival in pediatric ALF (44%-83%) vs single modality (approximately 30%-60%), with rapid ammonia clearance and HE resolution. In adults, however, adding TPE to CRRT shows no benefit and may even worsen outcomes post-transplant[62,75-79]. ACLF: Modest short-term survival gain (48%-65% at 28-90 days) over single therapy, mainly in HBV-related ACLF, with better MELD/SOFA reduction and sepsis control. In non-HBV and general ACLF-AKI cohorts, combined therapy is equivalent or inferior to TPE or CRRT alone, providing no consistent added survival advantage[42,72,75,80,81]

TPE: Therapeutic plasma exchange; CRRT: Continuous renal replacement therapy; DAMPs: Damage-associated molecular patterns; PAMPs: Pathogen-associated molecular patterns; HMGB1: High-mobility group box 1; LPS: Lipopolysaccharide; IL-6: Interleukin-6; CVVH: Continuous veno-venous hemofiltration; CVVHD: Continuous veno-venous hemodialysis; AKI: Acute kidney injury; ALF: Acute liver failure; INR: International normalized ratio; HE: Hepatic encephalopathy; ACLF: Acute-on-chronic liver failure; MELD: Model for end-stage liver disease; HBV: Hepatitis B virus; SOFA: Sequential organ failure assessment; HV-TPE: High-volume therapeutic plasma exchange; SV-TPE: Standard-volume therapeutic plasma exchange; DPMAS: Double plasma molecular adsorption system; RCA: Regional citrate anticoagulation; ECLS: Extracorporeal life support; ICU: Intensive care unit; TNF-α: Tumour necrosis factor-alpha; ICP: Intracranial pressure; RR: Risk ratio; MOF: Multiorgan failure; LT: Liver transplantation; SLED: Sustained low-efficiency dialysis; MAP: Mean arterial pressure; NGAL: Neutrophil gelatinase-associated lipocalin; CLIF: Chronic liver failure; iCa: Ionized calcium; TFS: Transplant-free survival; RCT: Randomised controlled trial; CI: Confidence interval; PLEX: Plasma exchange.

Recommendations (evidence-based framework)

Based on the data above, the pathophysiology of ALF and ACLF, along with outcomes using TPE/CRRT/combined/sequential treatments, Figure 3 depicts a flowchart that offers a clear, evidence-based framework with specific thresholds, regimens, and contraindications. It promotes clinical precision and adapts to resource limitations.

Figure 3 Proposed flowchart designed to guide clinicians in the precise use of therapeutic plasma exchange and continuous renal replacement therapy for managing acute liver failure and acute-on-chronic liver failure. The flowchart outlines decision-making pathways for patient selection, therapy initiation, sequencing (sequential or combined therapeutic plasma exchange-continuous renal replacement therapy), regimen selection, and monitoring, based on evidence from 2016-2025 (PubMed, EMBASE, Cochrane) and standardised guidelines (e.g., American Society for Apheresis, Asian Pacific Association for the Study of the Liver). ALF: Acute liver failure; ACLF: Acute-on-chronic liver failure; TPE: Therapeutic plasma exchange; CRRT: Continuous renal replacement therapy; HE: Hepatic encephalopathy; INR: International normalised ratio; AIH: Autoimmune hepatitis; HV: High volume; SV: Standard volume; AKI: Acute kidney injury; CVVH: Continuous veno-venous hemofiltration; CVVHDF: Continuous veno-venous hemodiafiltration; RCA: Regional citrate anticoagulation; MELD: Model for end-stage liver disease; HBV: Hepatitis B virus; FFP: Fresh frozen plasma.

CONCLUSION

In cases of ALF and acute-on-chronic liver failure, TPE and CRRT are no longer optional rescue measures; they are crucial, time-sensitive elements of modern management. TPE acts as the first-line treatment for hyperbilirubinaemia, coagulopathy, and inflammation in ALF (transplant-free survival increases by 10%-15%)[16] and ACLF grade 2-3 (90-day response rate 0.81)[40], with high-volume exchange preferred in ALF and standard-volume in ACLF. However, its risks of bleeding and infection (5%-15%) and its high cost require careful monitoring of INR, bilirubin, and cytokines[16,18,40,45]. CRRT is vital for stage 3 AKI, hyperammonemia, and fluid overload in both ALF (transplant-free survival relative risk 0.73)[64] and ACLF (90-day transplant-free survival 6%-23%)[71-74], ideally using CVVHDF with RCA anticoagulation to minimise citrate toxicity (12%) and bleeding (22%), with monitoring every 4-6 hours for ammonia and electrolytes[19,71,72,82]. The combination of TPE and CRRT (preferably sequential for safety), despite providing only modest survival benefits in particular groups, may be reserved for fulminant ALF or ACLF-3 with multiorgan failure (survival increase of 15%-25% compared to monotherapy)[75-81], offering synergistic clearance but with increased risks of fluid overload and citrate accumulation (increase of 10%-15%)[62,72,75]. This necessitates hybrid monitoring and specialised expertise to balance potential harms in select high-risk cases. The challenge is real, but the route is now clearly defined. Emerging biomarkers, such as IL-6 levels exceeding 100 pg/mL and NGAL levels above 400 ng/mL, can aid in guiding personalised therapy. Notably, IL-6 has been shown to predict response to TPE in 70% of inflammatory ALF cases; however, further validation across diverse populations is needed[22,23,26].

Future research should aim to develop cost-effective, automated systems, standardise treatment protocols, and reduce disparities in healthcare access. Clinical trials like CRITICAL (NCT 06987604) are exploring the effects of high-dose CRRT and early intervention. Moreover, scalable solutions such as portable TPE-CRRT devices and telemedicine for protocol dissemination are vital for improving survival and quality of life worldwide[30]. Integrating advanced diagnostic tools, including real-time cytokine profiling and point-of-care NGAL testing, could enhance the precision of therapies, especially in low-resource settings, thereby ensuring equitable access to these life-saving treatments.