ALSS is a crucial treatment modality for liver failure and plays a role in improving short-term prognosis. Currently, the optimal timing, dosing, frequency, and duration of the three fundamental techniques for ALSS treatment (CRRT, PE, and plasma adsorption) remain unclear [2]. CRRT could significantly reduce blood ammonia, intracranial hypertension, and cerebral edema-related mortality in patients with liver failure, potentially significantly enhancing the 21-day transplant-free survival [27, 28]. CRRT should be considered for patients with liver failure and grade III/IV hepatic encephalopathy, or when CRRT is indicated [2]. The therapeutic dosing of CRRT is the effluent volume with a recommended range of 20–25 ml/(kg·h), termed as delivered dose; when the anticipated CRRT duration is less than 24 h, it is advised to increase the prescribed therapeutic dosing, such as 25–30 ml/(kg·h) or higher, to achieve the delivered dose and fulfill therapeutic purpose [29,30,31]. The therapeutic dosing of PE is the volume of plasma replaced. Recent randomized controlled trials demonstrated that PE with 1–3 plasma volume could significantly improve short-term prognosis in patients with liver failure [32,33,34]. Therefore, PE is considered essential for ALSS treatment in liver failure [2]. The recommended therapeutic dosing for PE treatment is 1–1.5 plasma volume [3, 29, 35], potentially reaching up to three plasma volumes [35]. Based on evidence-based research, there are relatively clear recommendations regarding therapeutic dosing for CRRT and PE treatment. However, there is a lack of sufficient evidence to establish precise therapeutic dosing for plasma adsorption.

The therapeutic dosing of DPMAS treatment, the volume of processed plasma in the perfusion device, can be expressed in three ways: (1) indirectly hinted: therapeutic dosing (mL) = plasma separation rate (mL/h) × treatment duration (h); (2) direct representation, for example, 4500 mL; and (3) a combination of indirect hint and direct representation, which is the most comprehensive method. The DPMAS product manual suggests a therapeutic dosing of 3600–5400 mL with a treatment duration of 2–3 h. The “Standard Operating Procedures (SOP) for Blood Purification (2021 edition)” proposes a therapeutic dosing for plasma adsorption as 2–3 plasma volume with a treatment duration of 2–3 h [29]; While the “Expert consensus on clinical application of artificial liver and blood purification (2022 edition)” suggests setting the therapeutic dosing for plasma adsorption at a minimum of 1.2 plasma volume, generally within 2–3 plasma volume, with a minimum treatment duration of 2 h [3]. Nevertheless, there are notable discrepancies between the reported therapeutic dosing of DPMAS treatment in the literature and these recommendations, with large variations (Supplementary file 1: Table S1) [4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20]. Major studies have reported therapeutic dosing directly displayed as 3000–6000 mL [4, 7,8,9,10,11,12,13,14,15,16,17,18], while indirectly hinted therapeutic dosing are indicated as plasma separation rates of 20–50 mL/min with a treatment duration of 2–4 h [4,5,6, 10, 11, 18,19,20]. The normal human plasma volume is approximately calculated as follows: plasma volume (mL) ≈ 65 × ideal body weight (kg) × (1-hematocrit) [29], approximately 4% of the ideal body weight (40 ml/kg). Although the reported therapeutic dosing of DPMAS treatment is generally above 1.2 plasma volume, some therapeutic dosing is below two plasma volumes. This might be related to substantial variations in the treatment efficacy of DPMAS treatment; the immediate reduction ratios of total bilirubin ranged from 30% to 65% pre- and post-treatment (Supplementary file 1: Table S1). This prospective observational cohort study aimed to determine the appropriate therapeutic dosing for DPMAS treatment. Setting the plasma separation rate at 25 mL/min, the total volume of processed plasma for a 3-h session was 4500 mL. The study found that the total bilirubin levels and their reduction ratios in all patients, or patients with total bilirubin <425 μmol/L or ≥425 μmol/L changed gradually and significantly at four different time and different volume of processed plasma of DPMAS treatment (at 0.0 h (0 mL); at 2.0 h (3000 mL); at 2.5 h (3750 mL); at 3.0 h (4500 mL)) (all adjusted P for pairwise comparisons <0.001 and all adjusted P for trend <0.001), with no rebound in total bilirubin levels observed. These results suggest that at a plasma separation rate of 25 mL/min and a volume of processed plasma of 4500 mL over 3 h, the adsorption capacity of DPMAS remains unsaturated. It is appropriate to continue increasing the volume of the processed plasma of DPMAS treatment.

The DPMAS device has a limited capacity for adsorption, targeting bilirubin in the bloodstream. Once blood bilirubin levels decrease to a certain level, a dynamic equilibrium is reached, manifesting as sustained stability of total bilirubin levels during continuous DPMAS treatment. Given the large apparent distribution volume of bilirubin [36, 37], as blood bilirubin levels decrease, extravascular bilirubin gradually enters the bloodstream. Simultaneously, bilirubin originating from the liver continuously enters the blood stream. Although the apparent blood total bilirubin levels may not decrease, they actually reflect the removal of extravascular and liver-derived bilirubin by DPMAS treatment. Therefore, considering the absence of a statistical inflection point indicating an increase in total bilirubin levels, it is advisable to further increase the volume of processed plasma in DPMAS treatment to eliminate more bilirubin. Jung et al. investigated the capability of the molecular adsorbent recirculating system (MARS) and the fractionated plasma separation and adsorption system (FPSA) in eliminating bilirubin in severe liver disease over a 6-h treatment period [37]. They observed, from the perspective of bilirubin sources and distribution, that assessing treatment efficiency using pre- and post-treatment measurements of blood bilirubin levels significantly underestimated the actual capacity of MARS and FPSA to eliminate bilirubin (48% ± 10% vs. 54% ± 13%, P < 0.05) [37]. Because the double absorption technique is also used in MARS and FPSA, the findings from MARS and FPSA support our aforementioned inference during DPMAS treatment. When a patient’s blood total bilirubin levels remain stable without an inflection point indicating an increase, it is permissible to continue increasing the volume of processed plasma of DPMAS treatment to further eliminate bilirubin.

Bilirubin binds mainly to albumin in the blood and is transported to the liver for metabolism. The amount and function of albumin affects the metabolism of bilirubin [38, 39]. In this study, we found that the positive relationship between the volume of processed plasma and the reduction ratios of total bilirubin was less remarkable in patients with lower albumin levels (adjusted P for interaction = 0.017). This finding suggests that increasing albumin levels might help in bilirubin clearance during DPMAS treatment. We also found that the positive relationship between the volume of processed plasma and the reduction ratios of total bilirubin was less remarkable in patients with higher height (adjusted P for interaction = 0.027). A possible reason for this phenomenon is that the relative therapeutic dosing was smaller in patients with a higher height due to their larger plasma volume [29]. This finding suggests that the volume of processed plasma for DPMAS treatment in patients with a higher height should be increased.

DPMAS, PE, and post-dilutional continuous veno-venous hemofiltration (post-CVVH), a CRRT method, share similar technical principles and extracorporeal circuit connection methods. The key difference lies in post-CVVH using a hemofilter, discarding separated waste fluid, and replacing it with an equal volume of replacement fluid (with no dehydration) or less than the waste fluid volume of replacement fluid (with dehydration). PE employs a plasma separator, discards separated plasma, and replenishes it with an equal volume of allogeneic plasma (having no dehydration), akin to allogeneic plasma serving as the replacement fluid in post-CVVH without dehydration. Similarly, DPMAS utilizes a plasma separator, where separated plasma is adsorbed and processed before it is reintroduced into the body (without dehydration), akin to purified autologous plasma serving as the replacement fluid in post-CVVH without dehydration [40]. Consequently, during DPMAS and PE treatments, the fundamental principles of setting the CRRT parameters should be followed. For instance, maintaining an appropriate blood concentration ratio (≤20–25%) and filtration fraction (≤25–30%) to minimize significant blood concentration and blood protein-membrane reactions in the plasma separator area, thereby reducing the patient’s blood cell loss [29,30,31]. In clinical practice, it is more suitable to adjust the blood concentration ratio and control the filtration fraction based on the patient’s hematocrit, with the aim of increasing the unit-time volume of processed plasma and the duration to augment the therapeutic dosing of DPMAS treatment (Supplementary file 1: Table S2).

This study had several limitations. (1) Although total bilirubin levels represent the most significant parameter for change during ALSS treatment, bilirubin reduction is not the sole target of DPMAS treatment. The clearance of inflammatory mediators and other factors is equally crucial and should be considered as an assessment criterion for therapeutic dosing. (2) The sample size was based on a moderate effect; therefore, increasing the effect size might yield different results in studies with larger sample sizes. (3) Under the same volume of processed plasma, variations in the plasma separation flow and treatment duration could potentially yield different outcomes. (4) During DPMAS treatment, the change in total bilirubin is the most significant; thus, it is suitable to use it as a surrogate biomarker to find a more appropriate volume of processed plasma in DPMAS treatment. Several studies have focused on the impact of DPMAS plus PE treatment on patient outcomes and have achieved positive results [10, 41, 42]; however, high-quality evidence is still needed. The hypothesis that different volumes of processed plasma for DPMAS treatment may lead to different patient outcomes needs to be tested in the future.

In conclusion, the volume of processed plasma of DPMAS treatment could be more than 4500 mL. While ensuring patient safety, continued efforts could be made to increase the volume of processed plasma in order to enhance treatment efficacy, especially for patients with higher height or lower albumin levels who might require a higher VPP to achieve sufficient therapeutic efficacy. Future prospective multicenter cohort studies are warranted to further elucidate the appropriate therapeutic dosing for DPMAS treatment. Specifically, it is necessary to explore other potential biomarkers or factors in addition to total bilirubin levels. For instance, the role of patient individual characteristics, disease state, combined treatment regimens should be investigated. This approach would facilitate a more comprehensive understanding of the treatment response, thereby enabling the development of personalized treatment strategies. Such studies would provide high-quality evidence to precisely implement DPMAS treatment and improve the prognosis of patients with critically ill liver disease.