- Department of Critical Care Medicine, Union Hospital, Tongji Medical College, Huazhong University of Science and Technology, Wuhan, China
Therapeutic plasma exchange (TPE) is widely used in various fields of clinical practice. There are two main modes of TPE: centrifugal TPE (cTPE) and membrane TPE (mTPE). No matter which mode of TPE, anticoagulant therapy is essential to prevent clotting in the extracorporeal circuit. Therefore, selecting a suitable and safe anticoagulant is crucial. Currently, heparin and citrate are the most commonly used anticoagulants. Different anticoagulant regimens have varying advantages, disadvantages, and associated complications, making them suitable for different TPE modes. This review aims to summarize relevant research findings to provide clinicians with evidence to guide their selection of appropriate anticoagulants during TPE.
1 Introduction
Therapeutic plasma exchange (TPE) has been extensively applied in various fields, including hepatology, nephrology, hematology, neurology, immunology, and organ transplantation (1, 2). In clinical practice, there are two modes of TPE: centrifugal TPE (cTPE) and membrane TPE (mTPE) (3). Globally, the selection of TPE modes hinges on local economics and preferences (4). It has been reported that mTPE is prevalently utilized in China, Japan, and Germany, while cTPE is widely used in the United States (5, 6).
Anticoagulant therapy is important to prevent extracorporeal circuit clotting irrespective of the TPE mode. Frequent clotting in the circuit not only reduces the efficacy of treatment but also escalates the treatment cost and exacerbates blood loss. Conversely, excessive anticoagulation may lead to bleeding, posing a potential threat to the patients’ lives. Hence, it is imperative to choose a suitable and safe anticoagulant. Currently, heparin and citrate are the most commonly used anticoagulants. Nevertheless, the selection of anticoagulants for different TPE modes has been controversial (7–10). This study endeavors to summarize relevant research findings, aiming to offer evidence for clinicians to select appropriate anticoagulants during TPE.
2 Importance of anticoagulant therapy and coagulation status evaluation
2.1 Anticoagulant therapy
TPE involves separating the plasma of patients from whole blood using plasma separation technology. Subsequently, the plasma or plasma substitute products are replenished to clear pathogenic factors, such as autoantibodies, immune complexes, lipoproteins, and toxins in the plasma, mitigating the adverse effects of these factors and alleviating pathological processes (11). The blood clotting mechanism is activated once the blood is drawn. To ensure the smooth progress of treatment, anticoagulation therapy is recommended. Some studies have reported that at most treatment centers, TPE is performed without anticoagulation, and the incidence of tube clotting is 2.1%. Thus, anticoagulation therapy was considered non-mandatory. However, this recommendation was based on single-center work experience (12). Scholars from various countries recognize the importance of anticoagulants during TPE (3–11). Yuan et al. (7) reported a high clotting frequency in the filter in the non-anticoagulant group (33%), which increased the treatment duration and cost, and also contributed to the failure of treatment completion and the need to replace the pipeline. Procedural delay can have adverse effects on the plasma because it has a limited shelf life. Moreover, pipeline clotting can lead to anemia or thrombocytopenia in patients. Therefore, in clinical practice, anticoagulants are used for TPE. Failing to use anticoagulants cannot guarantee clinical therapeutic effects, can cause physical, mental, and economic damage to patients, and can increase the burden of nursing work.
2.2 Evaluation of the blood coagulation status
The use of anticoagulants raises the risk of bleeding. Therefore, patients need to be evaluated for bleeding and clotting status before TPE. This evaluation includes inquiring about patients’ past medical histories, assessing the risk of hemorrhagic and thromboembolic diseases, and evaluating the coagulation indices of patients (13). Evaluating coagulation status can assist clinicians in selecting the appropriate anticoagulation method and monitoring the effect of anticoagulants. For example, activated partial thromboplastin time (APTT), activated clotting time (ACT), prothrombin time (PT), thrombin time (TT), fibrinogen degradation products (FDP), fibrinogen and thromboela-stogram (TEG) are used to detect potential bleeding tendencies, while platelet count is used to determine platelet function. Consequently, clinicians must select appropriate anticoagulants and administer precise anticoagulant dosages according to patients’ coagulation function, bleeding risk, as well as the indications and contraindications.
3 Anticoagulant drugs, dosage, and monitoring
3.1 Systemic anticoagulation
3.1.1 Unfractionated heparin
UFH [molecular weight: 5,000–30,000 Dalton (Da)] is a commonly used anticoagulant in blood purification therapy. The anticoagulant effect of UFH can be attributed to the upregulation of antithrombin III (ATIII) activity and the inhibition of thrombin (IIA and Xa) activity, prolonging APTT (13–18). Systemic anticoagulant effects can be achieved in 3–5 min post-anticoagulant administration. Heparin is mainly metabolized by the liver and reticuloendothelial system (16). The physiological half-life of UFH is 30 min, which can be extended to 3 h in patients with renal failure (16, 17). At the beginning of TPE, patients are administered with UFH at a dose of 62.5–125 U/kg (0.5–1 mg/kg), and subsequently, UFH is administered at a dose of 1,250–2,500 U/h (10–20 mg/h) as needed. Dosing is stopped 30 min before the end of treatment (11). APTT is a reliable indicator of the anticoagulant effect and safety of heparin. The APTT value should be maintained at levels that are 1.5–2.5 times the physiological value to achieve anticoagulant effects and mitigate bleeding risk (14, 18). The advantages of UFH are its cost-effectiveness, ease of use, ease of monitoring the anticoagulant effect, and ability to be neutralized using protamine. The disadvantages of UFH include the unpredictable pharmacokinetics resulting in dosing variability, heparin resistance due to low antithrombin levels, the development of heparin-induced thrombocytopenia (HIT), and an increased risk of haemorrhage (13, 14). Previous studies have reported that the incidence rate of heparin-induced bleeding is 10–50%, which is caused by prolonged APTT. Therefore, heparin is contraindicated in patients who are bleeding or at high risk of bleeding due to recent surgery, trauma, and coagulation dysfunction, as well as in patients with HIT (Table 1).
3.1.2 Low-molecular-weight heparin
LMWH (molecular weight: approximately 4,000–6,000 Da) is produced through the chemical degradation or enzymatic hydrolysis of ordinary heparin (13, 15). LMWH exerts anticoagulant effects mainly by inhibiting Xa activation (13–18). The prolonged half-life of LMWH (2–4 h) is due to its clearance by the kidney (13, 14, 18). Patients with renal insufficiency and undergoing dialysis exhibit delayed LMWH metabolism. The excessive use of LMWH increases the risk of bleeding. In TPE, LMWH is intravenously administered at a dose of 60–80 IU/kg body weight 20–30 min before treatment without additional dosing. The anticoagulation effect of LMWH is monitored based on the anti-Xa levels with target levels being 0.25–0.35 IU/mL(14, 18). Compared with UFH, LMWH has the advantages of simple pharmacokinetics and reliable anticoagulant response. The disadvantages of LMWH include prolonged anticoagulation effects in patients with renal failure and poor reversal with protamine. Additionally, the need to test anti-Xa activity increases the monitoring cost. Similar to UFH, LMWH is contraindicated in patients at a high risk of bleeding due to its systemic effects (13, 14, 18).
3.1.3 Thrombin antagonists
Thrombin inhibitors (argatroban and bivalirudin) are used as an alternative to UFH in patients with HIT. Argatroban, a commonly used thrombin inhibitor in patients with HIT (14), is a synthetic small-molecule drug (molecular weight: 527 Da). Additionally, argatroban can reversibly bind to the active site of thrombin and inhibit thrombin-catalyzed reactions, including the formation of fibrin, the activation of coagulation factors V, VIII, and PC, and platelet aggregation (13). Argatroban is primarily metabolized by the liver (13, 14, 18, 19). The half-life of argatroban in individuals with physiological renal function is 15–30 min (13, 14, 18, 19). During TPE, patients are administered argatroban at a dose of 0.05–0.1 mg/kg and subsequently at a dose of 1–3 mg/h as needed. Dosing is stopped 20–30 min before the end of treatment.
Bivalirudin can be used as an alternative to argatroban in patients with liver and kidney failure. Bivalirudin has a short half-life, exhibits reversible thrombin binding, and is cleared via extra-renal and extra-hepatic mechanisms (20). The pharmacological advantages of bivalirudin, a hirudin analog, include decreased dependence on renal clearance and a low allergic reaction rate. The target APTT level is 1.5–2.0 times the physiological level of APTT (13, 19).
3.1.4 Nafamostat mesilate (NM)
NM, a serine protease inhibitor (molecular weight: 540 Da), mainly inhibits thrombin, Factor XIa, Xa, XII, the kallikrein-kinin system, the fibrinolysis system, the complement, and platelet activation (13). The liver and blood are the main sites for NM metabolism. The half-life of NM is 5–8 min, which has little effect on blood coagulation function in vivo. Thus, NM is a novel anticoagulant used in cardiopulmonary bypass, especially for patients at risk of bleeding or with active bleeding (13, 21). During TPE, 40 mg of NM in 1,000 mL saline is used for pipeline precharging. NM is administered at an initial dosage of 20–50 mg/h for continuous intravenous infusion (22). It is recommended to monitor the ACT or the APTT to guide the adjustment of NM dose. The dose is adjusted according to the risk of blood clotting, the anticoagulant effect, and the blood clotting function (22). Nakamura et al. (23) reported that NM exhibited good safety and efficacy profiles as an anticoagulant in TPE for high-risk bleeding populations. However, NM is associated with several side effects and is expensive (24, 25), which limits its clinical application.
3.2 Regional anticoagulation
3.2.1 Regional citrate anticoagulation (RCA)
The relative molecular weight of sodium citrate (citrate) is 294 Da. RCA is a special mode of in vitro anticoagulation therapy. The basic principle of RCA involves the injection of citrate into the extraction end of circulation loop. Citrate reversibly chelates ionized calcium (iCa) in the circuit. As a result, the coagulation mechanism is inhibited when the concentration of iCa in the circulation loop is less than 0.45 mmol/L. Previous studies have demonstrated that regulating in vitro iCa concentrations at 0.25–0.45 mmol/L can achieve the anticoagulation effect of the circulation loop (13, 14, 18, 26). However, an equal amount of cleared iCa should be added at the end of the circulation loop to ensure that the in vivo iCa level is 1.0–1.2 mmol/L (27, 28) (Table 1).
3.3 Non-anticoagulation therapy
Performing TPE without anticoagulants is a controversial approach. Currently, non-anticoagulant therapy is recommended for patients at high risk of bleeding and those with contraindications to citrate. Previous studies have reported that the incidence of clotting in the TPE extracorporeal circulation in the non-anticoagulation mode ranges from 6.3 to 33% (6, 7, 9). The incidence of clotting in non-coagulation therapy varies. Even when the non-anticoagulant mode does not increase the bleeding risk, the clotting system can be activated in some patients with high-risk bleeding and active bleeding. Additionally, the risk of clotting in the extracorporeal circulation pipeline increases significantly, which adversely affects the treatment efficiency and increases the medical burden on patients. However, non-anticoagulant mode is the preferred option for TPE in patients with contraindications to anticoagulants.
4 Selection of anticoagulants for different TPE modes
Heparin and citrate are the most commonly used anticoagulants. For patients with contraindications for both citrate and heparin, argatroban and NM can be used to prevent coagulation in the extracorporeal loop (29). Previously, citrate and heparin were commonly used for cTPE and mTPE, respectively. However, citrate has also been recently used for mTPE (8–10), especially for patients at high risk of bleeding. According to the World Apheresis Registry (WAR) (12), TPE is the most commonly used treatment for therapeutic apheresis, with cTPE and mTPE accounting for 34.8 and 0.8% of treatments, respectively. The most commonly used anticoagulation method is RCA.
4.1 Selection of anticoagulant for mTPE
Heparin anticoagulation (HA) is the most widely used method in mTPE and has the advantages of satisfactory anticoagulation effect, rapid onset, and low cost (7). In mTPE, approximately 70% of citrate enters the body because of the low filter clearance rate of citrate, which increases the metabolic burden and may induce complications such as metabolic alkalosis and hypocalcemia. Hence, several hemodialysis centers use HA as the standard anticoagulation procedure for mTPE (29) (Figure 1). However, clinical studies have reported significant side effects of HA(30). Heparin cannot be used in patients with HIT or severe bleeding. Therefore, researchers advocate the extensive application of RCA in mTPE. Recently, several hospitals have adopted RCA as the preferred anticoagulation method for mTPE (7, 31). The dose of citrate pumped into patients is directly correlated with its effects. Thus, several studies have focused on examining the minimum citrate required to meet the anticoagulant effect. TPE requires large infusions of fresh frozen plasma (FFP), the concentration of citrate in FFP ranges from 17 to 21 mmol/L (32). Therefore, compared to continuous renal replacement therapy (CRRT), TPE has a lower citrate demand. Yuan et al. (7) reported that during mTPE, the initial pumping rate of 4% citrate is set at 1.8% of the blood flow rate can effectively exert anticoagulation effects. Patients did not develop citrate-related side effects and electrolyte disturbances because of careful patient monitoring and intravenous calcium supplementation. This suggests that low-dose RCA is an effective anticoagulation method for mTPE. Additionally, calcium infusion should be provided at the end of the circuit of mTPE to maintain the physiological iCa level in vivo. Yuan et al. (7) suggested the initiation of 10% calcium gluconate at a rate of 25–30 mL/h with a monitoring interval of once every 30 min. If the parameters are stable, the monitoring interval can be extended appropriately. Thus, clinical studies have reported that the routine use of low-dose RCA during mTPE is safe and effective and can be widely used in clinical practice.

Figure 1. Anticoagulation regimen for membrane TPE. In membrane TPE, the patient’ plasma is separated through a hollow fiber filter. HA is used as the anticoagulation method. The replacement fluid (e.g., fresh frozen plasma or albumin) is infused post-filtration before returning to the patient. Membrane TPE, membrane therapeutic plasma exchange; HA, heparin anticoagulation.
4.2 Selection of anticoagulant for cTPE
Clinical studies have recommended that RCA is the optimal anticoagulation method for cTPE (8, 33). The clearance rate of citrate in cTPE can reach up to 80%. Hence, the incidence of citrate-induced toxic reactions and metabolic alkalosis is markedly mitigated (33). When citrate is used as an anticoagulant during cTPE, the ratio of the 4% citrate rate to the blood flow rate is set at 1.2:1, which can ensure effective anticoagulation in the cTPE pipeline (34) (Figure 2). However, the blood iCa concentration is markedly decreased because of the high demand for citrate and the high iCa clearance rate in cTPE. In patients using FFP as a replacement fluid, the citrate in the plasma further binds to the iCa in the blood. Therefore, continuous calcium supplementation is needed to prevent hypocalcemia during cTPE. Coirier et al. (33) revealed that infusing 10% calcium gluconate at a rate of 20 mL/h during cTPE, along with the anticoagulant effect of citrate, can prevent the occurrence of hypocalcemia.

Figure 2. Anticoagulation regimen for centrifugal TPE. In centrifugal TPE, blood is withdrawn from the patient, and RCA is used as the anticoagulation method. The blood undergoes centrifugation to separate plasma, which is then replaced with a replacement fluid (e.g., fresh frozen plasma or albumin) before being reinfused into the patient. Centrifugal TPE, centrifugal therapeutic plasma exchange; RCA, regional citrate anticoagulation.
5 Complications related to the use of TPE anticoagulants
5.1 Bleeding
Bleeding is one of the most common complications of anticoagulation therapy, especially when HA is used (35). HA exerts a systemic anticoagulant effect by prolonging APTT and increases the risk of bleeding. Previous studies have confirmed HA increases the risk of bleeding by 50% with every 10 s increase in APTT (36). Thus, heparin should not be administered to patients with heparin-induced HIT or AT-III deficiency (35–37). Weiss et al. (38) reported that citrate affected the activation and adhesion of platelets and white blood cells in a concentration-dependent manner, exerting anticoagulant effects. However, the low-dose infusion of citrate has enabled the maintenance of low concentrations of citrate in patients, minimally affecting the patients’ coagulation system and preventing bleeding complications (7). Several studies have demonstrated that RCA-TPE is associated with a low risk of bleeding (7, 39, 40). Jiao et al. (35) performed a multivariate analysis of bleeding events and identified HA as one of the independent risk factors for the occurrence of bleeding events in mTPE. Additionally, the authors revealed that the bleeding event in the HA-mTPE mode was 2.666 times higher than that in the RCA-mTPE mode (35). Therefore, the coagulation index and systemic bleeding status of patients should be carefully monitored, and the heparin dosage should be adjusted accordingly in HA-TPE mode. The use of HA in patients at a high risk of bleeding undergoing TPE is controversial (Table 2). If excessive anticoagulant use is detected, we must prepare the corresponding antagonists: for heparin and LMWH, protamine can be used as an antidote, while for citrate anticoagulants, increasing calcium infusion can help prevent bleeding caused by hypocalcemia.
5.2 HIT
UFH and LMWH can induce thrombocytopenia. A recent meta-analysis revealed that the risk rates of HIT with the usage of UFH and LMWH were 2.6 and 0.2%, respectively (41). Type I HIT is a non-immunogenic reaction that occurs 1–2 days after heparin administration. Patients with type I HIT exhibit a slight decrease in the platelet count without thrombus formation or bleeding, and this condition can be alleviated without stopping the drug. Type II HIT is an immunoreactive thrombocytopenia mediated by the combination of heparin and platelet factor IV to form a complex (PF4-H), which stimulates the production of IgG antibodies and activates platelets and the clotting pathway. Type II HIT occurs 5–10 days after heparin administration but also can occur within 24 h–3 weeks. Most cases of type II HIT are characterized by platelet reduction by more than 50% relative to the baseline value and can be accompanied by severe thromboembolism and acute systemic reactions (18, 41). Regular heparin and LMWH should be discontinued immediately upon diagnosis of HIT. These patients must be administered with argatroban or NM as anticoagulants.
5.3 Filter and pipeline coagulation
Filter and pipeline clotting, which is common in mTPE, occurs when the dose of anticoagulant is insufficient. Previous studies have reported that the average coagulation incidence in the HA-mTPE extracorporeal circulation pipeline is 1.8–7.5% (7, 9, 39, 42). Meanwhile, the incidence of coagulation in the RCA group is 0–4.4% (35). Patients with complex and changeable conditions and individual differences may have a high risk of bleeding along with blood hypercoagulability. Conventional doses of anticoagulants may not reach the effective concentration for anticoagulation, increasing the risk of blood clotting in the circulation loop. Therefore, before treatment, the patient’s condition, relevant coagulation indicators, and appropriate anticoagulant selection should be considered. The anticoagulant dose should be increased for patients with blood hypercoagulability and hyperlipidemia. Additionally, sufficient anticoagulant drugs should be prepared to avoid interruption of anticoagulant therapy. Moreover, during treatment, the coagulation status in the filters and pipeline must be monitored. When clotting occurs, the dosage of anticoagulant drugs can be appropriately increased to prevent unnecessary clotting events in the filters and pipeline.
5.4 Metabolic alkalosis
The concentration of citrate in the FFP is reported to be 17–21 mmol/L (32). RCA-TPE will additionally increase citrate concentration. In mTPE, 30% of the citrate is removed by the filter, while in cTPE, 80% of the citrate is removed by the filter (11). Citrate that is not removed during TPE enters the body and is metabolized to produce a large amount of bicarbonate. Unlike CRRT, TPE does not use an acidic replacement solution to neutralize the bicarbonate produced through citrate metabolism (7). Therefore, when citrate is used in TPE, the occurrence of metabolic alkalosis, especially in the mTPE mode, must be monitored. Metabolic alkalosis can occur during RCA-mTPE (7, 39, 40). Shunkwiler et al. (43) confirmed that RCA-mTPE was associated with a higher incidence of metabolic alkalosis than HA-mTPE. Metabolic alkalosis is likely to occur during continuous RCA-mTPE (35). The occurrence of metabolic alkalosis in RCA-cTPE has not been previously reported, which can be due to the limited number of clinical studies on cTPE, the high clearance rate of citrate in cTPE, and the low production of bicarbonate. However, further studies are needed to investigate the occurrence of metabolic alkalosis during RCA-cTPE. Therefore, the acid-base balance of patients undergoing RCA-TPE should be closely monitored to prevent the occurrence of metabolic alkalosis. Metabolic alkalosis, a complication of citrate use, can be easily corrected (44). For patients at risk of developing metabolic alkalosis, the pumping rate of citrate can be reduced without compromising the therapeutic effect. Severe metabolic alkalosis, which is common in patients with severe renal failure, can be corrected using CRRT (45). Kaushik et al. (46) revealed that the parallel or sequential combination of TPE and CRRT is technically feasible and safe for patients undergoing TPE and requiring hemodialysis at the same time and can effectively alleviate citrate-induced metabolic alkalosis (Table 2).
5.5 Citrate accumulation (CA)
CA is a serious complication of RCA. In patients with liver failure, severe shock, or hypoxemia, citrate that enters the body cannot be metabolized, which increases the incidence of CA (47, 48). Severe CA can lead to hypocalcemia and metabolic acidosis, which can be life-threatening for critically ill patients. The total calcium (TCa)/iCa ratio of ≥ 2.5 was used as an indirect standard to determine the occurrence of CA (48, 49). However, continuous citrate pumping in RCA-TPE increases the occurrence of hypocalcemia, raising the TCa/iCa ratio to ≥ 2.5 even without CA. The occurrence of CA should be comprehensively evaluated according to the following four criteria: serum Ca2+ level, TCa/iCa ratio, acid-base status, and anion gap change. These criteria have been widely used in the safety assessment of RCA-CRRT (35, 50). Based on Khadzhynov’s rigorous evaluation criteria, several studies have reported a low incidence of metabolic acidosis after RCA-mTPE (39, 40). The occurrence of CA has not been reported in patients undergoing RCA-cTPE, which may be related to the high rate of citrate clearance in the cTPE mode. However, further studies are needed for clinical verification (Table 2).
5.6 Hypocalcemia
Hypocalcemia (with or without symptoms) is one of the most common adverse effects of RCA, with incidence rates of 1.5–9% among patients undergoing TPE (51). The incidence of hypocalcemia is high due to the high clearance rate in the cTPE mode. Coirier et al. (33) revealed that the incidence of hypocalcemia during RCA-cTPE was 17.8–21.4% and that the incidence of hypocalcemia was higher when the replacement fluid was plasma. Hypocalcemia was classified according to Lee’s classification (37), and it was mainly grade I asymptomatic (73%) (33). Therefore, prophylactic calcium pump therapy and continuous iCa monitoring have been recommended for patients undergoing RCA-TPE (52). The risk of symptomatic hypocalcemia can be mitigated by supplementing calcium gluconate or calcium chloride (53). Numbness around the mouth and fingers, metallic taste in the mouth, nausea, chills, muscle cramps, and prolonged QT interval on the electrocardiogram are indicators of hypocalcemia. Communication disorders may lead to serious complications, such as hand-foot convulsions, arrhythmias, and seizures in patients with consciousness disorder, patients undergoing sedation and analgesia treatment, and pediatric patients. Therefore, iCa levels must be carefully monitored in these patients (51) (Table 2).
5.7 Other electrolyte imbalance
During RCA-TPE, the chelation of citrate with iCa increases intracellular sodium (iNa) levels in the body (1 mmol of 4% sodium citrate solution contains 3 mmol of iNa). However, it is reported that the occurrence of hypernatremia is rare after RCA-TPE (35). Citrate can also chelate magnesium. The removal of the citrate-magnesium complex by an extracorporeal circuit increases the risk of hypomagnesemia (40). However, hypomagnesemia cannot be detected using blood gas analysis. Thus, necessity of daily monitoring of magnesium levels in patients after TPE is debatable. An international survey of pediatric patients undergoing TPE revealed that 19% of respondents used magnesium supplements (38). However, magnesium supplementation is considered unnecessary for TPE (43).
6 Anticoagulation options for special patients
6.1 Anticoagulation regimen for patients at high risk of bleeding
The selection of the anticoagulation mode is challenging for critically ill patients undergoing TPE, especially for those who are at a high risk of bleeding due to surgery/trauma, active bleeding, and coagulation disorders. RCA, non-anticoagulation, and NM anticoagulation regimens have been used in patients at high risk of bleeding undergoing mTPE (7, 9, 25, 35). For critically ill patients with a high risk of bleeding, anticoagulation programs are selected based on the assessment of the patients’ coagulation function and bleeding risk, as well as anticoagulation indications and contraindications. HA and RCA are the two most widely used anticoagulant regimens in clinical practice. The use of HA or RCA in patients at high risk of bleeding undergoing TPE is controversial. Jiao et al. (35) revealed that the therapeutic effects of HA-TPE and RCA-TPE were similar, with no significant differences in the incidence of blood clotting during extracorporeal circulation. However, compared with HA, RCA can effectively reduce the risk of bleeding. Additionally, the blood transfusion requirement in the HA group was significantly higher than that in the RCA group, which may also be related to the increased number of bleeding events. Yuan et al. (7) demonstrated that the platelet and coagulation functions were not significantly different between RCA and non-anticoagulant groups among patients with bleeding tendencies. However, the incidence of pipeline coagulation in the non-anticoagulant group was higher than that in the RCA group. RCA is an effective and safe anticoagulation method for patients at high risk of bleeding. However, compared with HA-mTPE, RCA-mTPE has a higher risk of metabolic alkalosis and hypocalcemia during treatment, which requires careful monitoring and timely adjustment (35). Previous studies have reported the safety and effectiveness of NM anticoagulation in plasma exchange therapy in patients at high risk of bleeding (26). Clinicians can apply NM anticoagulation in plasma exchange therapy for patients with a high risk of bleeding.
6.2 Anticoagulation regimen for patients with liver failure
TPE has been used for the treatment of liver failure (54, 55). The EASL (56) Clinical Practical Guidelines on the management of acute (fulminant) liver failure cite high-volume TPE as the first-line treatment for acute liver failure. However, the use of anticoagulants for TPE in patients with liver failure is controversial because citrate is mainly metabolized by the liver. In patients with liver failure, impaired liver metabolism may lead to an increase in risk of CA, which is often exacerbated by renal impairment (57). Compared with RCA, HA can aggravate coagulation disorders in patients with liver failure, thereby increasing the risk of bleeding. Yuan et al. (7) reported that the incidence of bleeding in patients with liver failure undergoing HA-mTPE was 13.7–26.9%. Yuan et al. (42) suggested that non-anticoagulation therapy could reduce the risk of bleeding in patients with liver failure and that citrate in the plasma could meet the demand for anticoagulation. Therefore, the self-coagulation function and bleeding risk of the patients must be evaluated to select the appropriate anticoagulation regimen according to indications and contraindications. Recent studies have confirmed that patients with liver failure can metabolize citrate. Citrate used in CRRT has been confirmed to be safe for patients with liver failure (39, 58–60). Therefore, citrate is suggested as the first choice of anticoagulation for TPE in patients with liver failure (38, 57). Ma et al. (39) demonstrated that although the incidence of CA in patients with liver failure at 1 h post-TPE was 29.6%, citrate returned to physiological levels the next morning. Additionally, citrate was well tolerated without causing other discomforts. The common complications of RCA-TPE in patients with liver failure are hypocalcemia and metabolic acidosis (39). Therefore, RCA-TPE is recommended for patients with liver failure. However, the occurrence of metabolic acidosis and hypocalcemia in these patients must be monitored. Additionally, NM anticoagulation can be used for patients with contraindications to citrate, such as those with severe liver failure (13).
6.3 Anticoagulation regimen for patients receiving anticoagulant therapy
Inpatients or outpatients may receive anticoagulant therapy because of an underlying disease. For example, patients may receive UFH or LMWH to prevent venous thromboembolism and warfarin or direct oral anticoagulant (DOAC) for atrial fibrillation (43, 61). The use of anticoagulants for TPE in patients already receiving anticoagulant therapy varies among different centers. TPE may be performed after anticoagulants are no longer at therapeutic levels or have been reversed by antagonists (43, 61). When heparin and LMWH are used, the duration of TPE can be determined according to the half-life of the drug (62, 63). This is challenging for patients receiving warfarin or DOAC because these drugs has a long half-life, the availability of reversing agents is low, and the ability to easily monitor bleeding risk is poor. In individuals treated with warfarin, albumin is recommended as a replacement fluid. Alternatively, the anticoagulant medication can be switched to UFH or LMWH before treatment to control the anticoagulant effect (61, 64). Previous studies have reported (61–64) that TPE can be performed with citrate and heparin following standard protocols in patients who have received previous anticoagulant therapy. However, citrate and heparin increase the risk of bleeding complications in these patients.
7 Conclusion
Clinical TPE is mainly divided into cTPE and mTPE modes. RCA is recommended as the preferred anticoagulation method for both modes. For patients at high bleeding risk or with diseases such as liver failure, clinical studies have supported the use of RCA. However, the dosage of citrate must be strictly monitored because of the low clearance rate of citrate in the mTPE mode. Additionally, the incidence of CA and other related complications is relatively high. The most common complications of TPE using citrate anticoagulation are metabolic alkalosis and hypocalcemia. Although the symptoms are not serious, arterial blood gas, acid-base balance, and electrolyte levels must be monitored when citrate anticoagulation is applied. The occurrence of severe hypocalcemia can be clinically prevented through prophylactic calcium supplementation.
Author contributions
RY: Conceptualization, Data curation, Formal Analysis, Writing – original draft. HH: Conceptualization, Formal Analysis, Project administration, Resources, Supervision, Validation, Visualization. Writing – review & editing. YM: Conceptualization, Project administration, Investigation, Methodology, Writing – original draft. HG: Project administration, Resources, Software, Supervision, Validation, Writing – original draft. LM: Project administration, Validation, Methodology, Writing – original draft. SW: Validation, Investigation, Supervision, Writing – original draft. JW: Methodology, Project administration, Writing – original draft. JZ: Project administration, Supervision, Writing – original draft. XW: Investigation, Methodology, Writing – original draft. YW: Methodology, Formal Analysis, Writing – original draft. YY: Formal Analysis, Validation, Writing – original draft. TW: Validation, Supervision, Writing – original draft. RL: Validation, Conceptualization, Data curation, Formal Analysis, Funding acquisition, Project administration, Software, Visualization, Writing – review & editing.
Funding
The author(s) declare that financial support was received for the research and/or publication of this article. The study was supported by the National Natural Science Foundation of China (grant numbers 82002099, 82472224).
Acknowledgments
RL thank her friend Ms. Ting Zhao (JTM CCT), whose music and company have given her powerful spiritual support and encouragement.
Conflict of interest
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
The reviewer XR declared a shared parent affiliation with the authors to the handling editor at the time of review.
Publisher’s note
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.
Abbreviations
TPE, therapeutic plasma exchange; cTPE, centrifugal therapeutic plasma exchange; mTPE, membrane therapeutic plasma exchange; APTT, activated partial thromboplastin time; ACT, activated clotting time; PT, prothrombin time; TT, thrombin time; FDP, fibrinogen degradation products; TEG, thromboela-stogram; UFH, unfractionated heparin; ATIII, antithrombin III; Da, Dalton; HIT, heparin-induced thrombocytopenia; LMWH, low-molecular weight heparin; NM, Nafamostatmesilate; RCA, regional citrate anticoagulation; WAR, World Apheresis Registry; HA, heparin anticoagulation; CRRT, continuous renal replacement therapy; CA, citrate accumulation; TCa/iCa, total calcium/ionic calcium; DOAC, direct oral anticoagulant; FFP, fresh frozen plasma.
References
1. Connelly-Smith L, Alquist C, Aqui N, Hofmann J, Klingel R, Onwuemene O, et al. Guidelines on the use of therapeutic apheresis in clinical practice - Evidence-based approach from the writing committee of the American society for apheresis: The ninth special issue. J Clin Apher. (2023) 38:277–8. doi: 10.1002/jca.22043
2. Abe T, Matsuo H, Abe R, Abe S, Asada H, Ashida A, et al. The Japanese society for apheresis clinical practice guideline for therapeutic apheresis. Ther Apher Dial. (2021) 25:728–876. doi: 10.1111/1744-9987.13749
3. Chen Y, Sun X, Huang W, He F, Zhang C. Therapeutic apheresis in kidney diseases: An updated review. Ren Fail. (2022) 44:842–57. doi: 10.1080/0886022X.2022.2073892
4. Malchesky P, Koo A, Skibinski C, Hadsell A, Rybicki L. Apheresis technologies and clinical applications: The 2007 International apheresis registry. Ther Apher Dial. (2010) 14:52–73. doi: 10.1111/j.1744-9987.2009.00716.x
5. Córdoba J, Larrarte C, Medina M. Experience in therapeutic plasma exchange by membrane filtration at an academic center in Colombia: Registry of the first 500 sessions. J Clin Apher. (2015) 30:347–52. doi: 10.1002/jca.21391
6. Gashti C, Andreoli D, Patel D. Membrane-based therapeutic plasma exchange (mTPE): Technical and clinical experience. J Clin Apher. (2018) 33:38–45. doi: 10.1002/jca.21561
7. Yuan F, Li Z, Li X, Liu H. Application of regional citrate anticoagulation in membrane therapeutic plasma exchange. Int Urol Nephrol. (2020) 52:2379–84. doi: 10.1007/s11255-020-02581-0
8. Hafer C, Golla P, Gericke M, Eden G, Beutel G, Schmidt J, et al. Membrane versus centrifuge-based therapeutic plasma exchange: A randomized prospective crossover study. Int Urol Nephrol. (2016) 48:133–8. doi: 10.1007/s11255-015-1137-3
9. Brunetta Gavranić B, Bašić-Jukić N, Premužić V, Kes P. Membrane therapeutic plasma exchange with and without heparin anticoagulation. J Clin Apher. (2017) 32:479–85. doi: 10.1002/jca.21544
10. Kissling S, Legallais C, Pruijm M, Teta D, Vogt B, Burnier M, et al. A new prescription model for regional citrate anticoagulation in therapeutic plasma exchanges. BMC Nephrol. (2017) 18:84. doi: 10.1186/s12882-017-0494-9
11. Cervantes C, Bloch E, Sperati C. Therapeutic plasma exchange: Core curriculum 2023. Am J Kidney Dis. (2023) 81:475–92. doi: 10.1053/j.ajkd.2022.10.017
12. Mantilla J, Peña D, Molano-Triviño A, Zúñiga E, Castellanos-De La Hoz JC. Divide et Vinces, therapeutic apheresis in nephrological clinical practice. Blood Purif. (2024) 53:162–9. doi: 10.1159/000534455
13. Expert Consensus Working Group. Chinese expert consensus on the application of anticoagulant technology in renal replacement therapy for critical patients. Chinese J Nephrol. (2023) 39:155–64. doi: 10.3760/cma.j.cn441217-20220809-00815
14. Legrand M, Tolwani A. Anticoagulation strategies in continuous renal replacement therapy. Semin Dial. (2021) 34:416–22. doi: 10.1111/sdi.12959
15. Hirsh J, Warkentin T, Shaughnessy S, Anand S, Halperin J, Raschke R, et al. Heparin and low-molecular-weight heparin: Mechanisms of action, pharmacokinetics, dosing, monitoring, efficacy, and safety. Chest. (2001) 119:64S–94S. doi: 10.1378/chest.119.1_suppl.64s
16. Brunet P, Simon N, Opris A, Faure V, Lorec-Penet A, Portugal H, et al. Pharmacodynamics of unfractionated heparin during and after a hemodialysis session. Am J Kidney Dis. (2008) 51:789–95. doi: 10.1053/j.ajkd.2007.12.040
17. Shen J, Winkelmayer W. Use and safety of unfractionated heparin for anticoagulation during maintenance hemodialysis. Am J Kidney Dis. (2012) 60:473–86. doi: 10.1053/j.ajkd.2012.03.017
18. Oudemans-van Straaten H, Wester J, de Pont A, Schetz M. Anticoagulation strategies in continuous renal replacement therapy: Can the choice be evidence based? Intensive Care Med. (2006) 32:188–202. doi: 10.1007/s00134-005-0044-y
19. Reddy B, Grossman E, Trevino S, Hursting M, Murray P. Argatroban anticoagulation in patients with heparin-induced thrombocytopenia requiring renal replacement therapy. Ann Pharmacother. (2005) 39:1601–5. doi: 10.1345/aph.1G033
20. Kiser T, MacLaren R, Fish D, Hassell K, Teitelbaum I. Bivalirudin versus unfractionated heparin for prevention of hemofilter occlusion during continuous renal replacement therapy. Pharmacotherapy. (2010) 30:1117–26. doi: 10.1592/phco.30.11.1117
21. Nakae H, Tajimi K. Pharmacokinetics of nafamostat mesilate during continuous hemodiafiltration with a polyacrylonitrile membrane. Ther Apher Dial. (2003) 7:483–5. doi: 10.1046/j.1526-0968.2003.00088.x
22. Hu J, Xie Y, Shen B, Xu J, Ding X, Mao Y. Society of nephrology, shanghai medical association, expert consensus on application of nafamostat mesilate in anticoagulation for blood purification writing group. expert consensus on application of nafamostat mesilate in anticoagulation for blood purification. SMJ. (2024) 47:129–44. doi: 10.19842/j.cnki.issn.0253-9934.2024.03.001
23. Nakamura Y, Tada K, Matsuta M, Murai A, Tsuchiya H. Anaphylactic reactions caused by nafamostat mesylate during hemodialysis before surgery for carpal tunnel syndrome. Case Rep Nephrol. (2021) 29:1148156. doi: 10.1155/2021/1148156
24. Han W, San Bok J, Cho H, Yu J, Na M, Kang S, et al. Single-center experience of extracorporeal membrane oxygenation mainly anticoagulated with nafamostat mesilate. J Thorac Dis. (2019) 11:2861–7. doi: 10.21037/jtd.2019.06.30
25. Schneider A, Joannes-Boyau O. Regional citrate anticoagulation for CRRT: Still hesitating? Anaesth Crit Care Pain Med. (2021) 40:100855. doi: 10.1016/j.accpm.2021.100855
26. Liu S, Xu S, Yin L, Yang T, Jin K, Zhang Q, et al. Emergency Medical Doctor Branch of the Chinese Medical Doctor Association. Management of regional citrate anticoagulation for continuous renal replacement therapy: Guideline recommendations from Chinese emergency medical doctor consensus. Mil Med Res. (2023) 10:23. doi: 10.1186/s40779-023-00457-9
27. Zarbock A, Küllmar M, Kindgen-Milles D, Wempe C, Gerss J, Brandenburger T, et al. Effect of regional citrate anticoagulation vs systemic heparin anticoagulation during continuous kidney replacement therapy on dialysis filter life span and mortality among critically ill patients with acute kidney injury: A randomized clinical trial. JAMA. (2020) 324:1629–39. doi: 10.1001/jama.2020.18618
28. Bauer P, Ostermann M, Russell L, Robba C, David S, Ferreyro B, et al. Plasma exchange in the intensive care unit: A narrative review. Intensive Care Med. (2022) 48:1382–96. doi: 10.1007/s00134-022-06793-z
29. Vrielink H, Le Poole K, Stegmayr B, Kielstein J, Berlin G, Ilhan O, et al. The world apheresis association registry, 2023 update. Transfus Apher Sci. (2023) 62:103831. doi: 10.1016/j.transci.2023.103831
30. Beurskens D, Huckriede J, Schrijver R, Hemker H, Reutelingsperger C, Nicolaes G. The anticoagulant and nonanticoagulant properties of heparin. Thromb Haemos. (2020) 120:1371–83. doi: 10.1055/s-0040-1715460
31. Halpin M, Chen B, Singer R. Efficacy, safety, and calcium balance of an accelerated citrate anticoagulated membrane-based plasma exchange algorithm. Blood Purif. (2022) 51:70–4. doi: 10.1159/000515827
32. Brandenburger T, Dimski T, Slowinski T, Kindgen-Milles D. Renal replacement therapy and anticoagulation. Best Pract Res Clin Anaesthesiol. (2017) 31:387–401. doi: 10.1016/j.bpa.2017.08.005
33. Coirier V, Lesouhaitier M, Reizine F, Painvin B, Quelven Q, Maamar A, et al. Tolerance and complications of therapeutic plasma exchange by centrifugation: A single center experience. J Clin Apher. (2022) 37:54–64. doi: 10.1002/jca.21950
34. Keklik M, Çelik S, Yıldızhan E. Comparison of centrifugal and membrane filtration modalities on therapeutic plasma exchange. J Clin Apher. (2022) 37:217–22. doi: 10.1002/jca.21961
35. Jiao J, Yu Y, Wei S, Tian X, Yang X, Feng S, et al. Heparin anticoagulation versus regional citrate anticoagulation for membrane therapeutic plasma exchange in patients with increased bleeding risk. Ren Fail. (2023) 45:2210691. doi: 10.1080/0886022X.2023.2210691
36. Garcia D, Baglin T, Weitz J, Samama M. Parenteral anticoagulants: Antithrombotic therapy and prevention of thrombosis, 9th ed: American college of chest physicians evidence-based clinical practice guidelines. Chest. (2012) 141:e24S–e43S. doi: 10.1378/chest.11-2291
37. Lee G, Arepally G. Anticoagulation techniques in apheresis: From heparin to citrate and beyond. J Clin Apher. (2012) 27:117–25. doi: 10.1002/jca.21222
38. Weiss R, Fischer M, Weber V. The impact of citrate concentration on adhesion of platelets and leukocytes to adsorbents in whole blood lipoprotein apheresis. J Clin Apher. (2017) 32:375–83. doi: 10.1002/jca.21519
39. Ma Y, Chen F, Xu Y, Wang M, Zhou T, Lu J, et al. Safety and efficacy of regional citrate anticoagulation during plasma adsorption plus plasma exchange therapy for patients with acute-on-chronic liver failure: A pilot study. Blood Purif. (2019) 48:223–32. doi: 10.1159/000500408
40. Peng X, Xie X, Yin J, Yu S, Zhang C, Zhao S. Anticoagulation effect and safety observation of regional citrate anticoagulation for double-filtration plasmapheresis in critical patients. Blood Purif. (2020) 49:542–9. doi: 10.1159/000505613
41. Martel N, Lee J, Wells P. Risk for heparin-induced thrombocytopenia with unfractionated and low-molecular-weight heparin thromboprophylaxis: A meta-analysis. Blood. (2005) 106:2710–5. doi: 10.1182/blood-2005-04-1546
42. Yuan S, Qian Y, Tan D, Mo D, Li X. Therapeutic plasma exchange: A prospective randomized trial to evaluate 2 strategies in patients with liver failure. Transfus Apher Sci. (2018) 57:253–8. doi: 10.1016/j.transci.2018.02.001
43. Shunkwiler S, Pham H, Wool G, Ipe T, Fang D, Biller E, et al. Therapeutic Apheresis Subsection of the AABB. The management of anticoagulation in patients undergoing therapeutic plasma exchange: A concise review. J Clin Apher. (2018) 33:371–9. doi: 10.1002/jca.21592
44. Schneider A, Journois D, Rimmelé T. Complications of regional citrate anticoagulation: Accumulation or overload? Crit Care. (2017) 21:281. doi: 10.1186/s13054-017-1880-1
45. Sanchez A, Balogun R. Therapeutic plasma exchange in the critically Ill patient: Technology and indications. Adv Chronic Kidney Dis. (2021) 28:59–73. doi: 10.1053/j.ackd.2021.03.005
46. Kaushik M, Liew Z, Sewa D, Phua G, Cao L, Krishnamoorthy T, et al. Description of parallel and sequential configurations for concurrent therapeutic plasma exchange and continuous kidney replacement therapy in adults. J Clin Apher. (2021) 36:211–8. doi: 10.1002/jca.21854
47. Delaney M, Capocelli K, Eder A, Schneiderman J, Schwartz J, Sloan S, et al. An international survey of pediatric apheresis practice. J Clin Apher. (2014) 29:120–6. doi: 10.1002/jca.21301
48. Oudemans-van Straaten H, Ostermann M. Bench-to-bedside review: Citrate for continuous renal replacement therapy, from science to practice. Crit Care. (2012) 16:249. doi: 10.1186/cc11645
49. Tolwani A, Wille K. Advances in continuous renal replacement therapy: Citrate anticoagulation update. Blood Purif. (2012) 34:88–93. doi: 10.1159/000342378
50. Khadzhynov D, Schelter C, Lieker I, Mika A, Staeck O, Neumayer H, et al. Incidence and outcome of metabolic disarrangements consistent with citrate accumulation in critically ill patients undergoing continuous venovenous hemodialysis with regional citrate anticoagulation. J Crit Care. (2014) 29:265–71. doi: 10.1016/j.jcrc.2013.10.015
51. Mokrzycki M, Balogun R. Therapeutic apheresis: A review of complications and recommendations for prevention and management. J Clin Apher. (2011) 26:243–8. doi: 10.1002/jca.20303
52. Gashti C. Membrane-based therapeutic plasma exchange: A new frontier for nephrologists. Semin Dial. (2016) 29:382–90. doi: 10.1111/sdi.12506
53. Kankirawatana S, Huang S, Marques M. Continuous infusion of calcium gluconate in 5% albumin is safe and prevents most hypocalcemic reactions during therapeutic plasma exchange. J Clin Apher. (2007) 22:265–9. doi: 10.1002/jca.20142
54. Goel A, Zachariah U, Daniel D, Eapen C. Growing evidence for survival benefit with plasma exchange to treat liver failure. J Clin Exp Hepatol. (2023) 13:1061–73. doi: 10.1016/j.jceh.2023.06.002
55. Mhanna M, Srour O, Nawras M, Mentrose J, Assaly R, Kubal C, et al. Plasma exchange for acute and acute-on-chronic liver failure: A systematic review and meta-analysis. Liver Transpl. (2024) 30:127–41. doi: 10.1097/LVT.0000000000000231
56. European Association for the Study of the Liver, Electronic address:ZWFzbG9mZmljZUBlYXNsb2ZmaWNlLmV1; Clinical practice guidelines panel, Wendon J, Panel members Cordoba J, et al. EASL clinical practical guidelines on the management of acute (fulminant) liver failure. J Hepatol. (2017) 66:1047–81. doi: 10.1016/j.jhep.2016.12.003
57. Chris-Olaiya A, Kapoor A, Ricci K, Lindenmeyer C. Therapeutic plasma exchange in liver failure. World J Hepatol. (2021) 13:904–15. doi: 10.4254/wjh.v13.i8.904
58. Klingele M, Stadler T, Fliser D, Speer T, Groesdonk H, Raddatz A. Long-term continuous renal replacement therapy and anticoagulation with citrate in critically ill patients with severe liver dysfunction. Crit Care. (2017) 21:294. doi: 10.1186/s13054-017-1870-3
59. Zhang W, Bai M, Yu Y, Li L, Zhao L, Sun S, et al. Safety and efficacy of regional citrate anticoagulation for continuous renal replacement therapy in liver failure patients: A systematic review and meta-analysis. Crit Care. (2019) 23:22. doi: 10.1186/s13054-019-2317-9
60. Qi W, Liu J, Li A. Regional citrate anticoagulation or heparin anticoagulation for renal replacement therapy in patients with liver failure: A systematic review and meta-analysis. Clin Appl Thromb Hemost. (2023) 29:10760296231174001. doi: 10.1177/10760296231174001
61. Hodulik K, Root A, Ledbetter L, Onwuemene O. Effects of therapeutic plasma exchange on anticoagulants in patients receiving therapeutic anticoagulation: A systematic review. Transfusion. (2019) 59:1870–9. doi: 10.1111/trf.15191
62. Kaplan A, Raut P, Totoe G, Morgan S, Zantek N. Management of systemic unfractionated heparin anticoagulation during therapeutic plasma exchange. J Clin Apher. (2016) 31:507–15. doi: 10.1002/jca.21441
63. Schinzel H, Berghoff K, Beuermann I, Sauer O, von Mach M, Weilemann L. Anticoagulation with low-molecular-weight heparin (dalteparin) in plasmapheresis therapy: Initial experience. Transfusion. (2006) 46:624–9. doi: 10.1111/j.1537-2995.2006.00755.x
Keywords: therapeutic plasma exchange, centrifugal TPE, membrane TPE, anticoagulant therapy, anticoagulant drug
Citation: Yao R, Huang H, Ming Y, Guo H, Miao L, Wang S, Wan J, Zhou J, Wang X, Wu Y, Yuan Y, Wang T and Li R (2025) Anticoagulant regimens for different therapeutic plasma exchange modes. Front. Med. 12:1568333. doi: 10.3389/fmed.2025.1568333
Received: 29 January 2025; Accepted: 28 May 2025;
Published: 18 June 2025.
Edited by:
Eleni Gavriilaki, Aristotle University of Thessaloniki, GreeceReviewed by:
Xiao Ran, Huazhong University of Science and Technology, ChinaBo Hu, Zhongnan Hospital of Wuhan University, China
Copyright © 2025 Yao, Huang, Ming, Guo, Miao, Wang, Wan, Zhou, Wang, Wu, Yuan, Wang and Li. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
*Correspondence: Ruiting Li, NDk4Njc2NzcyQHFxLmNvbQ==
†These authors share first authorship