In the reviewed cases, PE was used as a secondary therapeutic option for patients who did not respond effectively to first-line treatments. Most patients were diagnosed with grade 3 or grade 4 CRS, and 9 patients developed neurotoxicity. All of them were treated with tocilizumab and corticosteroids before PE. The PE procedure involved the removal of plasma with subsequent replacement using fresh frozen plasma or albumin. PE was generally administered over a series of consecutive days, ranging from 1 to 4 times depending on the severity of CRS and the individual response. For instance, Xiao et al. reported that a single patient underwent PE for 3 consecutive days, which resulted in marked improvement in CRS symptoms and a reduction in cytokine levels (17)​.

Before PE treatment, patients exhibited elevated levels of inflammatory markers, particularly IL-6, TNF-α, and IL-10. After undergoing PE, there was a consistent decrease in these cytokines, which correlated with clinical improvements​. In the retrospective cohort study by Pu et al, cytokine levels such as IL-6 and CRP dropped significantly post-PE, indicating the efficacy of PE in alleviating CRS-related symptoms ​ (16). However, it still existed that several patients experienced the CRS progression and eventually died after PE. As shown in the table, the phenomenon that IL-6 levels remained increasing after PE was observed in all those 3 dead patients. This might be explained by the fact that the rate of cytokine removal by PE has fallen below the rate of cytokine production. Therefore, we propose that early intervention of PE might be an effective approach to treat these extremely severe CRS, this opinion will be discussed in the Discussion section.

Although preexisting studies predominantly focus on ALL, the efficacy of PE in treating CAR-T related CRS in NHL and MM was also definite. As shown in Table 2 , the reduction in IL-6 levels following PE treatment in four NHL patients (average 80%) was comparable to that observed in ALL patients (average 75%), while the future studies might need to expand the sample size of non-ALL cases to further validate the generalizability of PE.

PE is a blood purification therapy that involves removing the plasma of a patient and replacing it with fresh frozen plasma or other substitute fluids (22). The main objective of PE is to eliminate excessive cytokines, immune complexes, and other harmful substances that contribute to disease progression (23, 24). The use of PE in medicine could date back several decades, initially being used for autoimmune diseases such as systemic lupus erythematosus and myasthenia gravis (25, 26). Gradually, its application has expanded to treat acute kidney injury, toxicological emergencies and blood disorders (27, 28).

In recent years, with the rise of CAR-T cell therapy, PE has been increasingly applied for the management of severe CRS, particularly in patients with conventional treatments failing to control symptoms (16–18). In CAR-T cell therapy, PE is used to remove inflammatory cytokines, such as IL-6, TNF-α and other molecular cytokines that are released after the interaction between cancer cells and CAR-T cells.

Except for CRS caused by CAR-T cell therapy, PE has also been explored for managing CRS caused by other conditions, including viral infections such as COVID-19 (29). In cases of COVID-19, PE has been shown to reduce cytokine levels and improve clinical outcomes, particularly in patients experiencing severe inflammatory responses (30–32). A key difference between CAR-T-related CRS and COVID-19-induced CRS is the underlying immune activation: CAR-T cell therapy causes a rapid proliferation of engineered T-cells that directly attack cancer cells, triggering an inflammatory response; while COVID-19 related CRS results from the body’s response to viral infection, often leading to pulmonary damage and multi-organ failure (14, 33). This distinction in pathophysiology might influence the timing and volume of PE required. For example, PE should be administrated earlier when treating CAR-T related CRS, compared with infection related CRS, as CAR-T cells can exert their cytotoxic effects on tumor cells rapidly. Especially in treating hematologic malignancies, CAR-T cells can encounter and attack tumor cells immediately after being infused into peripheral blood, leading to the release of a large number of cytokines in the short time.

In addition to PE, other blood purification techniques like hemofiltration (HF) and hemodiafiltration (HDF) have been used for managing CRS. HF operates using convection to remove small molecules and excess fluid from the bloodstream, which is particularly useful for fluid overload or the removal of small molecular toxins, such as some inflammatory cytokines (34–36). However, HF is less effective in clearing large molecular weight cytokines, which limits its utility in the treatment of severe CRS (37). On the other hand, HDF combines the mechanisms of both diffusion and convection, which allows for more comprehensive clearance of small and medium-sized molecular weight cytokines (38, 39). HDF is often favored when a more extensive removal of inflammatory cytokines is needed. However, HDF requires more complex equipment and procedures, increasing its resource demands and operational challenges.

Each blood purification technique has its unique advantages and disadvantages in the management of CRS. PE is particularly advantageous for rapidly removing large molecular weight cytokines associated with severe CRS, especially in those patients who fail to respond to conventional therapies. This makes PE highly effective in improving symptoms such as fever, hypoxia and coagulopathy. Also, PE might be a feasible method to avoid or alleviate neurotoxicity, as it can directly reduce the damage to the blood-brain barrier caused by inflammatory cytokines. However, PE also has several drawbacks. Firstly, it is not effective in removing small molecule toxins and can be technically challenging, as it might require specialized equipment and multiple sessions of plasma replacement. Additionally, PE carries a risk of hemodynamic instability, which requires close monitoring during the procedure. Compared with PE, although HF is simpler and easier to perform, it is limited in its ability to clear large molecular cytokines (38). It is effective in addressing fluid overload and removing small molecular toxins, but its role in CRS treatment is less prominent. HDF combines the progress of both diffusion and convection, offering a more comprehensive solution for removing both small and medium molecular cytokines, making it a strong option for severe CRS. However, it is highly resource-intensive and requires high-end equipment, which might lead to severe complications such as hypotension and bleeding (40). As shown in Table 3 , we compare the advantages and disadvantages of these techniques, emphasizing their treatment mechanisms, efficacy, patient suitability and operational difficulties​.

To summarize, although PE is an effective method to remove various cytokines, the accessibility and cost of PE may limit its use due to the need for specialized equipment and expertise. In contrast, HF is less costly and simpler, but less effective in clearing large cytokines. HDF can balance both small and medium molecular toxin clearance, but it also requires complex equipment and may cause hemodynamic instability.

The timing of PE and the volume of plasma replacement are critical factors influencing the success of the treatment in CRS management. Nowadays, PE is most used when conventional therapies fail to control severe CRS. However, according to our experience, early intervention with PE might be able to prevent irreversible organ damage and neurotoxicity caused by continuous damage of cytokines. In our opinion, the timepoint to conduct PE is that the level of IL-6 exceeds 1000 pg/ml, or the increase of IL-6 is over 500 pg/ml within 24 hours. Typically, the volume of plasma exchanged is based on the patient’s body weight and hematocrit, with the following formula: PE Volume (L) = 0.065 × Body Weight (kg) × (1 − Hematocrit). However, for patients with severe CRS, additional sessions or a larger plasma volume may be required (41). The future direction of PE lies in optimizing treatment protocols, including determining the best timing, frequency of treatment and the appropriate volume of plasma to replace. Standardizing these protocols will provide more consistent clinical guidelines for the use of PE in CRS treatment. Furthermore, combination therapies involving PE and other immunomodulatory agents may increase the overall effectiveness while reducing the side effects of individual treatments. Advances in other blood purification technologies may lead to the development of more efficient and user-friendly devices, reducing the complexity of the procedure and making it more accessible and safer for severe CRS patients. Additionally, ongoing research into the exact mechanisms of CRS and how blood purification techniques can be combined with CAR-T cell therapy could significantly improve treatment outcomes in the future.