Secondary Logo

Journal Logo

Polymyxin B Direct Hemoperfusion Using Regional Citrate–Calcium Anticoagulation: A Case Report

Sidoti, Anna MD*; Brogi, Etrusca MD*; Morse, Joshua; Collareta, Michele MD*; Vetrugno, Luigi MD; Giunta, Francesco MD*; Forfori, Francesco MD*

doi: 10.1213/XAA.0000000000000392
Case Reports: Case Report

Direct hemoperfusion with polymyxin B (PMX-DHP) is an extracorporeal treatment to add to conventional therapy during unresponsive endotoxic septic shock. So far, only heparin has been used as an anticoagulant during polymyxin B therapy. We present a case report of a postsurgical septic patient treated with 2 cycles of PMX-DHP using citrate anticoagulation. Monitoring of serum calcium, postcartridge calcium, and acid–base balance was performed. The treatments were accomplished without complications. To our knowledge, this is the first published report on the use of citrate anticoagulation during PMX-DHP. We conclude that citrate anticoagulation is feasible during hemoperfusion therapy in patients with increased hemorrhagic risk.

From the *Department of Anesthesia and Intensive Care, University of Pisa, Pisa, Italy; McGill University, Montreal, Canada; and Department of Anesthesia and Intensive Care, University of Udine, Udine, Italy.

Accepted for publication June 3, 2016.

Funding: None.

The authors declare no conflicts of interest.

Address correspondence to Etrusca Brogi, MD, Department of Anesthesia and Intensive Care, University of Pisa, Via Paradisa 2, Pisa 56124, Italy. Address e-mail to

Direct hemoperfusion therapy with a polymyxin B (PMX-B) immobilized fiber cartridge (Toraymyxin column, Toray Industries Inc, Tokyo, Japan) is a well-tolerated and safe treatment for Gram-negative septic shock (also known as endotoxic shock) unresponsive to conventional therapy. The main mechanism of action is the direct adsorption of circulating lipopolysaccharide with PMX-B fixed on cartridge fibers.1 The PMX-B column has also been reported to remove inflammatory cells such as activated monocytes and neutrophils and, as a consequence, the exaggerated systemic immune response of patients can be reduced.2 Even if not in agreement with all the trials published so far,3,4 numerous clinical studies have shown that PMX-B hemoperfusion leads to an improvement in tissue oxygenation, in hemodynamic response, in renal function, and in decreasing overall mortality.5,6 Although awaiting the results of the EUPHRATES and the EUPHAS-2 trials to obtain more accurate results on the real efficacy of direct hemoperfusion with polymixyn B (PMX-DHP) treatments for endotoxic shock,7 we think that the use of PMX-DHP treatments in patients with septic shock is still justified.

Anticoagulation is essential for maintaining the fluidity of extravascular blood in the extracorporeal circuit. Both citrate and heparin are used as anticoagulants during continuous renal replacement therapy, but so far only heparin-based anticoagulation has been used in the setting of PMX-DHP therapy. Heparin anticoagulation is associated with much more bleeding complications and heparin-induced thrombocytopenia.8 Furthermore, sepsis is frequently associated with thrombocytopenia and hemostasis alterations; these changes could be exacerbated by heparin-induced thrombocytopenia.9 Moreover, postsurgical septic patients have an increased bleeding risk (ie, recent surgery, low platelet count, and low coagulation index on thromboelastography).10 Based on these data, Kidney Disease Improving Global Outcomes guidelines suggest using regional citrate anticoagulation, rather than heparin during continuous renal replacement therapy, in critically ill patients who do not have contraindications for citrate.11

Consequently, in light of the accumulating evidence in favor of regional citrate anticoagulation, and our direct experience with citrate anticoagulation therapy in different settings (continuous veno-venous hemodialysis [CVVHD], continuous veno-venous hemodialfiltration [CVVHDF], and coupled plasma filtration adsorption), we decided to use citrate anticoagulation for direct hemoperfusion with polymyxin B treatment.12

We report a case of a postsurgical septic patient with a high bleeding risk from thrombocytopenia, in which we performed 2 cycles of PMX-B hemoperfusion treatments using regional citrate as an anticoagulation agent.

Back to Top | Article Outline


A 71-year-old man was admitted to our intensive care unit for severe peritonitis caused by gastric leakage. Nine days before, he underwent a partial gastrectomy because of gastric cancer. Comorbidities included chronic lymphatic leukemia. The patient was admitted from the operating room, where he underwent an explorative laparotomy, surgical repair of gastric leakage, and abdominal toilette.

At the time of admission, he was in sinus tachycardia with a blood pressure of 75/55 mm Hg and a Spo2 of 90%, and he was oligoanuric (Table 1). He was sedated and intubated. Mechanical ventilation was started with low tidal volume (6 mL/kg), a positive-end expiratory pressure setting of 12 cm H2O, and a Fio2 of 60%. Arterial and central venous catheters were inserted, and fluid resuscitation was started. A fluid challenge with albumin (20 g/dL) and lactated Ringer solution was attempted. Because of the persistence of hypotension, norepinephrine was continuously administered (increased up to 0.45 μg/kg/min). The blood–gas analysis revealed mixed respiratory and metabolic acidosis (pH, 7.28; Pco2, 47 mm Hg; lactate, 3.7 mmol/L), the Pao2/Fio2 ratio was 103, and the central venous saturation was 62%.

Table 1.

Table 1.

A diagnosis of septic shock was made and blood cultures, bronchoalveolar lavage, and surveillance swabs were collected. This was followed by the administration of broad-spectrum antibiotics (ie, piperacillin/tazobactam 4.5-g bolus followed by continuous infusion of 18 g/d, gentamicin 320 mg/d, and metronidazole 500 mg/4 times a day) and antifungal drugs (fluconazole 600 mg/day). Because of the high dose of vasopressor and the patient’s negative response to the fluid challenge, hydrocortisone infusion was started (dose: 200 mg/d). We also performed the endotoxin activity assay that showed an endotoxin activity level of 0.65. Procalcitonin was also measured during admission and for the following days (Table 1).

Because of the severity of the clinical situation, we decided to begin PMX-B hemoperfusion within 6 hours from admission using an adaptation of the protocol used for CVVHD (Calcium-Citrate CVVHD Fresenius System).13 In fact, because of the increased risk of bleeding of this patient (recent surgery, low platelet count, and low coagulation index on thromboelastography), we decided to use citrate regional anticoagulation associated with a strict monitoring of systemic pH, bicarbonate, base excess, and both serum and postfilter calcium.14

We assembled the Toraymyxin filter column and set up a blood flow of 100 mL/min. Anticoagulant infusion was started in the preblood pump site using a solution of trisodium citrate 4% (citrate, 136 mmol/L); the citrate infusion was calculated to begin at a blood citrate concentration of 3.0 mmol/L (according to the formula shown in the Figure).



We started citrate infusion at 132 mL/h (total volume of citrate infusion during the first cycle: 264 mL) monitoring calcium concentration (systemic and postfilter). A central line was used for calcium chloride reinfusion (0.068 mmol/mL CaCl2—2 g in 50 mL NaCl 0.9%). Citrate infusion was then adapted to maintain a postfilter calcium concentration of approximately 0.35 mmol/L.14 A systemic calcium target was set to maintain a normal ionized calcium blood level.

The treatment lasted 2 hours. We repeated a second treatment 24 hours later using the same protocol (total volume of citrate infusion during the second cycle: 282 mL); the second treatment also lasted 2 hours. During the 2 cycles, there were no episodes of filter clotting, no bleeding, and no metabolic complications (ie, hypocalcemia/hypercalcemia or metabolic disorders). Data from the 2 cycles of PMX-B hemoperfusion treatments are shown in Table 2.

Table 2.

Table 2.

For the worsening of oliguria, continuous venous venous hemofiltration was started between the 2 cycles (blood flow, 150 mL/h; ultrafiltration rate, 2300 mL/h; fluid removal, 200 mL/h, full length of 24 hours, using Prismocitrate 10/2 as regional anticoagulation, Gambro Dasco S.p.A., Sondrio, Italy). On the first day after admission, we also started intravenous immunoglobulin therapy (IgM-enriched preparation, dose 5 mL/kg/d for 3 consecutive days).

Gas exchange and hemodynamic parameters gradually recovered. The SOFA score during admission was 18 points; 24 hours later, after the 2 PMX-DHP treatments, it decreased to 9 points. Consequently, the patient was successfully extubated on the fifth postoperative day. Kidney function parameters slowly improved; on the fourth postoperative day, renal replacement therapy was no longer necessary, and renal function fully recovered. Only the procalcitonin level showed a worsening until the third day from admission (procalcitonin on day 3 was 19.67 ng/dL): these conflicting data have scarce significance in the picture of the global improvement of hemodynamic and oxygenation parameters.

The patient was discharged from the intensive care unit to the subintensive care unit on the sixth day. At the time of the writing of this case report, the patient was in a surgical ward because of the formation of a duodenal fistula.

Back to Top | Article Outline


The 2 cycles of PMX-B hemoperfusion treatments using citrate as a regional anticoagulation agent were performed without complications. To monitor the efficacy and safety of regional anticoagulation in this setting, we needed to frequently assess blood calcium, pH, bicarbonate levels, and also postfilter calcium because there is no available PMX-B hemoperfusion circuit programmed to use citrate as an anticoagulant agent.

We decided to perform PMX-B hemoperfusion treatments with citrate in this patient because he was a postsurgical patient with septic shock with severe peritonitis. This type of patient is at high risk of bleeding complications. At the same time, the sepsis-mediated coagulation derangement led to thrombocytopenia and hemostasis alterations (both hypercoagulability and hypocoagulability).9 Taking into account that heparin anticoagulation is associated with a high risk of bleeding complications and heparin-induced thrombocytopenia, we believed that citrate, instead of heparin, represented a safer option for our patient.

To prevent filter clotting, we used a concentrated solution of citrate (trisodium citrate 4%). We chose this type of citrate formulation to obtain the maximum anticoagulation effect by administering the lowest volume of citrate. In fact, the treatment lasted only 2 hours; consequently, the infusion of just 2 mL/min of citrate was not able to create any predilution effect (thus reducing the efficacy of the filter). Even if the citrate was highly concentrated, for the short duration of the treatments (2 hours each), we administered small quantities of citrate solution (35 mmol). As a consequence, citrate and sodium load did not determine significant metabolic and electrolyte derangement even in our patient with altered citrate clearance (ie, sepsis). Furthermore, the constant monitoring of postfilter calcium concentration allowed us to accurately titrate the citrate infusion. Lastly, the citrate–calcium complex cannot be absorbed by the filter during the hemoperfusion treatment. As a result, the complex enters the systemic circulation where the citrate is metabolized in mitochondria, releasing the bound calcium. Subsequently, it is very important to accurately titrate the infusion of calcium to prevent systemic accumulation.

To our knowledge, this is the first time that citrate anticoagulation was used during a direct hemoperfusion treatment with PMX-B. Consequently, we did not know the effect of citrate on the performance of the Toraymyxin filter column in terms of biocompatibility and cartridge effectiveness. Moreover, we cannot predict whether the interaction between PMX-B and lipopolysaccharide is calcium-dependent. Studies in vitro and in animal models are needed to investigate the possible interaction between PMX-B and citrate because some questions remain to be answered. We acknowledge the limitations of this initial report and intend to examine the effect of citrate on the performance of the Toraymyxin filter column in further studies.

Back to Top | Article Outline


Our case report shows that citrate anticoagulation is feasible during PMX-DHP treatments in patients with increased hemorrhagic risk. Further studies are needed to compare the safety, effectiveness, and advantages of using regional citrate anticoagulation versus traditional heparin as well as the possible interaction between PMX-B and citrate during PMX-B hemoperfusion treatment.

Following the rules of our local ethics committee, written informed consent to participate and to publish this case report was obtained from the patient.

Back to Top | Article Outline


1. Ronco C, Piccinni P, Kellum J. Rationale of extracorporeal removal of endotoxin in sepsis: theory, timing and technique. Contrib Nephrol. 2010;167:2534.
2. Ronco C, Klein DJ. Polymyxin B hemoperfusion: a mechanistic perspective. Crit Care. 2014;18:309.
3. Iwagami M, Yasunaga H, Doi K, et al. Postoperative polymyxin B hemoperfusion and mortality in patients with abdominal septic shock: a propensity-matched analysis. Crit Care Med. 2014;42:11871193.
4. Payen DM, Guilhot J, Launey Y, et al.; ABDOMIX Group. Early use of polymyxin B hemoperfusion in patients with septic shock due to peritonitis: a multicenter randomized control trial. Intensive Care Med. 2015;41:975984.
5. Cruz DN, Antonelli M, Fumagalli R, et al. Early use of polymyxin B hemoperfusion in abdominal septic shock: the EUPHAS randomized controlled trial. JAMA. 2009;301:24452452.
6. Vincent JL, Laterre PF, Cohen J, et al. A pilot-controlled study of a polymyxin B-immobilized hemoperfusion cartridge in patients with severe sepsis secondary to intra-abdominal infection. Shock. 2005;23:400405.
7. Antonelli M, Ronco C. Polymyxin B hemoperfusion in sepsis: growing body of evidence and occasional conflicting results. Blood Purif. 2015;39:III.
8. Lee G, Arepally GM. Anticoagulation techniques in apheresis: from heparin to citrate and beyond. J Clin Apher. 2012;27:1171125.
9. Amaral A, Opal SM, Vincent JL. Coagulation in sepsis. Intensive Care Med. 2004;30:10321040.
10. Müller MC, Meijers JC, Vroom MB, Juffermans NP. Utility of thromboelastography and/or thromboelastometry in adults with sepsis: a systematic review. Crit Care. 2014;18:R30.
11. Kidney Disease: Improving Global Outcomes (KDIGO) Acute Kidney Injury Work Group. KDIGO Clinical Practice Guideline for Acute Kidney Injury. Kidney Int. 2012; 2: 89115.
12. Cubattoli L, Teruzzi M, Cormio M, Lampati L, Pesenti A. Citrate anticoagulation during CVVH in high risk bleeding patients. Int J Artif Organs. 2007;30:244252.
13. Available at: Accessed March 7, 2016..
Copyright © 2016 International Anesthesia Research Society