Portal de Periódicos da CAPES

The use of continuous renal replacement therapy in series with extracorporeal membrane oxygenation

2009; Elsevier BV; Volume: 76; Issue: 12 Linguagem: Inglês

10.1038/ki.2009.383

1523-1755

María José Santiago, Amelia Sánchez, Jesús López‐Herce, Rosario Pérez, Jimena del Castillo, Javier Urbano, Ángel Carrillo,

Antibiotics Pharmacokinetics and Efficacy

A large percentage of patients on extracorporeal membrane oxygenation (ECMO) require continuous renal replacement therapy (CRRT) usually performed through a different venous access or by introducing a filter into the ECMO circuit. Here, we evaluated the efficacy and safety of including a CRRT machine in the circuit by connecting its inlet line after the centrifugal pump and its outlet line before the oxygenator. We tested the function of the combined system initially in a closed circuit, followed by an experimental animal study, and, finally, in a clinical trial with six children. Both machines functioned adequately and there were no significant changes in the pressures of the ECMO circuit after the introduction of the CRRT device, thus achieving the preset negative balances and normalization of the serum urea and creatinine concentrations. The mean life of the filters was about 138 h, and only one filter needed changing due to clotting. Our study shows that the introduction of a CRRT device into the ECMO circuit is a safe and effective technique that improves fluid balance, increases filter life, and does not cause complications. For these reasons, this may be a good method for performing CRRT in patients on ECMO. A large percentage of patients on extracorporeal membrane oxygenation (ECMO) require continuous renal replacement therapy (CRRT) usually performed through a different venous access or by introducing a filter into the ECMO circuit. Here, we evaluated the efficacy and safety of including a CRRT machine in the circuit by connecting its inlet line after the centrifugal pump and its outlet line before the oxygenator. We tested the function of the combined system initially in a closed circuit, followed by an experimental animal study, and, finally, in a clinical trial with six children. Both machines functioned adequately and there were no significant changes in the pressures of the ECMO circuit after the introduction of the CRRT device, thus achieving the preset negative balances and normalization of the serum urea and creatinine concentrations. The mean life of the filters was about 138 h, and only one filter needed changing due to clotting. Our study shows that the introduction of a CRRT device into the ECMO circuit is a safe and effective technique that improves fluid balance, increases filter life, and does not cause complications. For these reasons, this may be a good method for performing CRRT in patients on ECMO. Extracorporeal membrane oxygenation (ECMO) is used for cardiac and/or respiratory support in patients in whom other measures have failed. When using this technique, up to 50% of patients require continuous renal replacement therapy (CRRT).1.Cavagnaro F. Kattan J. Godoy L. et al.Continuous renal replacement therapy in neonates and young infants during extracorporeal membrane oxygenation.Int J Artif Organs. 2007; 30: 220-226PubMed Google Scholar Renal failure, hypervolemia, and edema are frequent in patients on ECMO.2.Conrad S.A. Rycus P.T. Dalton H. Extracorporeal Life Support Registry Report 2004.ASAIO J. 2005; 51: 4-10Crossref PubMed Scopus (250) Google Scholar There are a number of methods for performing CRRT during ECMO. The most widely used is to introduce a hemofiltration filter into the ECMO circuit and to control the ultrafiltrate volume using an intravenous infusion pump. Another possibility is to perform renal replacement therapy through a venous access independent from the ECMO circuit. It is also possible to include a CRRT machine in the ECMO circuit. It has been in use in several ECMO departments and has been published as a single case,3.Hotti D.K. StGeorge-Hyslop C. Geary D. et al.Continuous renal replacement therapy (CRRT) in children using the AQUIARIUS.Nephrol Dial Transplant. 2006; 21: 2296-2300Crossref PubMed Scopus (14) Google Scholar but there are no studies that have analyzed the efficacy of this possibility. The objective of the present study has been to test the efficacy and safety of the inclusion of a CRRT machine in the ECMO circuit. The results are summarized in Table 1. Both machines functioned correctly and there were no significant changes in the pressures in the ECMO circuit. Initially, the Prisma device (Hospal, Spain) detected that the inlet pressure was positive and triggered the ‘arterial line disconnection’ alarm. However, this machine allows the alarm to be silenced, and the system continued functioning normally, achieving the preset negative balances.Table 1ECMO and CRRT variables in the closed circuit (in vitro) and in the experimental animal (in vivo) at different ECMO flowsIn vitroIn vivoStudyLow-flow ECMOHigh-flow ECMOMinimal assistanceMaximal assistanceECMO Blood flow (l/min)0.45–0.61–2.360.650.8 Preoxygenator pressure90–9287–144166–168160–176 Postoxygenator pressure71–7585–129167–157155–172 Suction pressure19–2218–21-15 to -24-23 to -26CRRT Blood flow (ml/min)40–18090–18040–10040–160 Ultrafiltrate flow (ml/h)100100–30000 Replacement flow (ml/h)100–350100–1000150–300150–1000 Dialysate flow (ml/h)01000600600–1000 Inlet pressure70–2588–20139–175107–170 Outlet pressure68–157105–213175–200175–236 Filter pressure66–203115–261185–230180–300 Transmembrane pressure0–110–6021–3317–49 Filter pressure drop0–4513–5131–660–4 Effluent pressure50505050CRRT, continuous renal replacement therapy; ECMO, extracorporeal membrane oxygenation.Pressure is measured in mm Hg.The pressures remained stable in the ECMO circuit throughout the experiment, and were not affected by the inclusion of CRRT. The inlet pressure to the hemofilter may be seen to be positive. Open table in a new tab CRRT, continuous renal replacement therapy; ECMO, extracorporeal membrane oxygenation. Pressure is measured in mm Hg. The pressures remained stable in the ECMO circuit throughout the experiment, and were not affected by the inclusion of CRRT. The inlet pressure to the hemofilter may be seen to be positive. The ECMO and CRRT parameters used, and the pressures and flows obtained are shown in Table 1. The system was used for 2 h with no incidences. Of the 23 patients treated using ECMO in our pediatric intensive care unit, 19 required CRRT. In two cases, the HF filter was introduced into the circuit, in 11, CRRT was performed through an independent venous access, and in six, the CRRT machine was connected to the ECMO circuit. These six patients aged between 3 days and 8 years, with body weights between 2.3 and 23 kg, were included in the study. The data, including number of filters used, mean life of the filters, heparinization, negative balances, and initial and final urea and creatinine concentrations, are shown in Table 2. The indication for CRRT was renal failure and/or hypervolemia.Table 2Patient dataPatient number12aPatient 2 was analyzed before and after heart transplant.2aPatient 2 was analyzed before and after heart transplant.3456Age (months)6969630.1247Weight7232362.997DiagnosisHLVS Glenn SurgeryHLVS Fontan SurgeryHeart transplantDilated cardiomyopathyHLVS Norwood SurgeryDilated cardiomyopathyHeart transplantIndicationHypervolemiaARFARFHypervolemiaHypervolemiaHypervolemiaARF+hypervolemiaTechniqueCVVHFCVVHDFCVVHDFCVVHFCVVHFCVVHFCVVHDFFilters0.6 m20.9 m20.9 m20.6 m20.6 m20.6 m20.6 m2Number of filters1212113Mean of filters life371162719969186191Heparin in CRRT+ECMO (U/kg/h)5+245+145+205+185+105+185+3CRRT Blood flow (ml/min)40909050406040 Ultrafiltrate flow (ml/h)151108030305030 Replacement fluid flow (ml/h)200500500150150200100 Dialysate flow (ml/h)09501000000300 Inlet pressure (mm Hg)-10102504513098137 Outlet pressure (mm Hg)12619725497153153156 Filter pressure (mm Hg)142252297116160182170 Filter pressure drop (mm Hg)18533720123112 Transmembrane pressure (mm Hg)391115225242732 Effluent pressure (mm Hg)>50>50>50>50>50>50>50Total negative balance (ml)56622,81932,30296712161883411,232Initial creatinine (mg/dl)0.410.840.60.250.670.450.36Final creatinine (mg/dl)0.190.780.590.320.20.360.2Initial urea (mg/dl)20524247363632Final urea (mg/dl)10455244155018ECMOBefore inclusion of CRRT Blood flow (l/min)0.541.81.90.960.420.640.95 Revolutions (rpm)2110251523751730214022402275 Premembrane pressure (mm Hg)10015910974122124122 Postmembrane pressure (mm Hg)1001359866119113116 Suction pressure-90-16-6-2-96After inclusion of CRRT Blood flow (l/min)0.511.81.90.960.410.640.94 Revolutions (rpm)2110251523751730214022402275 Premembrane pressure (mm Hg)10916612280122123124 Postmembrane pressure (mm Hg)10014410772117114118 Suction pressure-91-21-1-3-106ARF, acute renal failure; CRRT, continuous renal replacement therapy; CVVHDF, continuous veno-venous hemodiafiltration; CVVHF, continuous veno-venous hemofiltration; ECMO, extracorporeal membrane oxygenation; HLVS, hypoplastic left ventricle syndrome.Flows and pressures are represented by mean values.a Patient 2 was analyzed before and after heart transplant. Open table in a new tab ARF, acute renal failure; CRRT, continuous renal replacement therapy; CVVHDF, continuous veno-venous hemodiafiltration; CVVHF, continuous veno-venous hemofiltration; ECMO, extracorporeal membrane oxygenation; HLVS, hypoplastic left ventricle syndrome. Flows and pressures are represented by mean values. Both the ECMO and the CRRT functioned correctly in all cases, accurately achieving the preset negative balances, and with normalization of the urea and creatinine concentrations. There were no changes in the pressures in the ECMO circuit before or after inclusion of the CRRT machine. The entry pressure of the CRRT circuit was negative in only one patient; in the remainder, the pressure was positive, though this did not affect the functioning of the system. The patients tolerated the technique well, with no evidence of hypotension or other adverse effects. The mean life of the filters was 138.4 h (range, 37–271 h) (Table 2). In four patients, the same CRRT filter was used throughout the duration of ECMO. Only one filter had to be changed because of clotting. The rest of the filter changes were performed routinely because the patient had to be transferred to hemodynamic study. According to the Extracorporeal Life Support Organization,2.Conrad S.A. Rycus P.T. Dalton H. Extracorporeal Life Support Registry Report 2004.ASAIO J. 2005; 51: 4-10Crossref PubMed Scopus (250) Google Scholar 36% of patients on ECMO require CRRT, and this has reached 50% in other series.1.Cavagnaro F. Kattan J. Godoy L. et al.Continuous renal replacement therapy in neonates and young infants during extracorporeal membrane oxygenation.Int J Artif Organs. 2007; 30: 220-226PubMed Google Scholar Acute renal failure occurs in 50% of patients on ECMO.4.Kelly Jr, R.E. Phillips J.D. Foglia R.P. et al.Pulmonary edema and fluid mobilization as determinants of the duration of ECMO support.J Pediatr Surg. 1991; 26: 1016-1022Abstract Full Text PDF PubMed Scopus (56) Google Scholar,5.Roy B.J. Cornish J.D. Clark R.H. Venovenous extracorporeal membrane oxygenation affects renal function.Pediatrics. 1995; 95: 573-578PubMed Google Scholar The management of renal dysfunction and hypervolemia using CRRT in patients on ECMO improves fluid balance, resolves electrolyte imbalances, helps to improve calorie delivery by eliminating the need for fluid restriction, and reduces diuretic requirements.6.Hoover N.G. Heard M. Reid C. et al.Enhanced fluid management with continuous venovenous hemofiltration in pediatric respiratory failure patients receiving extracorporeal membrane oxygenation support.Intensive Care Med. 2008; 34: 2241-2247Crossref PubMed Scopus (89) Google Scholar In addition, the possibility for withdrawal of extracorporeal assistance is linked to the reduction of edema.4.Kelly Jr, R.E. Phillips J.D. Foglia R.P. et al.Pulmonary edema and fluid mobilization as determinants of the duration of ECMO support.J Pediatr Surg. 1991; 26: 1016-1022Abstract Full Text PDF PubMed Scopus (56) Google Scholar,5.Roy B.J. Cornish J.D. Clark R.H. Venovenous extracorporeal membrane oxygenation affects renal function.Pediatrics. 1995; 95: 573-578PubMed Google Scholar There are advantages and drawbacks to each of the three methods for performing CRRT in patients on ECMO. At first sight, it would appear simplest to perform CRRT through a venous access independent from the ECMO circuit. However, if the patient has not previously required CRRT, canalization of a new central venous line with a large-caliber catheter after starting ECMO is complicated, and anticoagulation increases the risks. The introduction of a hemofiltration filter into the ECMO circuit without a ‘in-line’ CRRT machine is the most widely used. The filter inlet is connected after the ECMO pump and the outlet is connected once again to the ECMO circuit; in circuits with a reservoir, the blood is usually returned to the reservoir (Figure 1). The effluent volume is controlled by an intravenous infusion pump included in the ultrafiltrate line, and this must be measured hourly. This technique has the advantage of being relatively simple and cheap.1.Cavagnaro F. Kattan J. Godoy L. et al.Continuous renal replacement therapy in neonates and young infants during extracorporeal membrane oxygenation.Int J Artif Organs. 2007; 30: 220-226PubMed Google Scholar,6.Hoover N.G. Heard M. Reid C. et al.Enhanced fluid management with continuous venovenous hemofiltration in pediatric respiratory failure patients receiving extracorporeal membrane oxygenation support.Intensive Care Med. 2008; 34: 2241-2247Crossref PubMed Scopus (89) Google Scholar However, there is no control over the blood flow through the hemofilter and intravenous infusion pumps are not accurate when used to slow down the passage of high-pressure ultrafiltrate fluid.7.Jenkins R. Harrison H. Chen B. et al.Accuracy of intravenous infusion pumps in continuous renal replacement therapies.ASAIO J. 1992; 38: 808-810PubMed Google Scholar We used a 0.3 m2 filter (Renaflow, Minntech, Minneapolis, USA) introduced into the ECMO circuit in two patients. The ultrafiltrate volume was controlled using intravenous infusion pumps (IVAC model 560, Alaris-Cardinal, San Diego, USA), detecting a large discrepancy between the preset ultrafiltrate volume and the volume obtained, due to the inability of the intravenous infusion pump to hold back the filtration sufficiently; this led to the need for very strict control by the nursing staff. Another drawback is that there is no monitoring of the pressures in the hemofiltration circuit, and clotting or rupture of the filter cannot be detected early. For these reasons, some authors propose filter changes at preset times,1.Cavagnaro F. Kattan J. Godoy L. et al.Continuous renal replacement therapy in neonates and young infants during extracorporeal membrane oxygenation.Int J Artif Organs. 2007; 30: 220-226PubMed Google Scholar although changes can lead to complications due to the need to manipulate the circuit. The possibility of including a venovenous CRRT machine in the ECMO circuit was proposed as a possible solution for these problems.8.Fernández C. López-Herce J. Flores J.C. et al.Prognosis in critically ill children requiring continuous renal replacement therapy.Pediatr Nephrol. 2005; 20: 1473-1477Crossref PubMed Scopus (50) Google Scholar Our study is the first to have analyzed the function of a CRRT machine in an ECMO circuit in vitro, in an animal model, and, finally, in children. The CRRT and ECMO functioned correctly, without complications, in all three phases of the study. In our experience, the inclusion of a renal filtration device in the ECMO circuit improves the monitoring of filter function and the accuracy of fluid balance. It has been reported that in some patients the real fluid removed by Prisma monitor was different from what was reported by the machine. The amount of variance was significant in some cases (relative to small size of the patient). For this reason, it could be necessary to measure the real fluid removed to make sure the accuracy of fluid balance. The blood flow in the CRRT machine is high and constant, avoiding alarms being triggered due to difficulty of blood extraction or return. In addition, CRRT can be withdrawn in a controlled manner, without affecting ECMO function, if the patient has to be transferred to hemodynamics or if clotting of the filter occurs. Owing to the stability of the system, the absence of problems of blood entry and return, and the dose of heparin received, filter life (138.4 h) was significantly longer than when CRRT is performed through an independent venous access (36.8 h).9.Del Castillo J. López-Herce J. Cidoncha E. et al.Circuit life span in critically ill children on continuous renal replacement treatment: a prospective observational evaluation study.Crit Care. 2008; 12: R93Crossref PubMed Scopus (47) Google Scholar Furthermore, the presence of accurate monitoring systems for circuit pressures and ultrafiltrate volume improves the control and safety of the patient, and reduces nursing workload. The theoretical drawbacks to this technique are that it adds complexity to the circuit and can increase costs.6.Hoover N.G. Heard M. Reid C. et al.Enhanced fluid management with continuous venovenous hemofiltration in pediatric respiratory failure patients receiving extracorporeal membrane oxygenation support.Intensive Care Med. 2008; 34: 2241-2247Crossref PubMed Scopus (89) Google Scholar We conclude that the inclusion of a CRRT device in the ECMO circuit is a safe and effective technique that improves fluid balance control, increases filter life, and does not lead to complications. For these reasons, it may be considered a good method for performing CRRT in patients on ECMO. The ECMO circuit used was a heparin-coated, closed circuit with no venous reservoir, and composed of a centrifugal pump (Jostra Rotaflow HL20, Maquet, Rastatt, Germany) and hollow fiber oxygenator (Quadrox, Jostra, Maquet, Rastatt, Germany). The total priming volume was 450 ml. CRRT was performed using the Prisma machine with polyacrylonitrile (AN69) membrane filters, with surface areas of 0.6 m2 and 0.9 m2 and total priming volumes of the circuits of 100 ml and 120 ml, respectively. Connection of the CRRT device into the ECMO circuit was performed in the following manner: the inlet (arterial) line of the CRRT circuit was connected after the centrifugal pump by a three-way tap that was also used for the infusion of heparin, and the outlet (venous) line was connected to the circuit at another tap before the oxygenator (Figure 2). The study was performed in three phases. First, the function of the two devices was tested in a closed circuit primed using colloids. We varied the ECMO blood flow and the CRRT parameters (flows of blood, replacement fluid, and dialysis solution), recording the pressures in the two devices. The system was then used in the experimental surgery laboratory on a minipig of 7 kg body weight, maintained with mechanical ventilation, propofol, fentanyl, and muscle relaxation with atracurium. The international guidelines for the care of experimental animals were observed. The external jugular vein and the carotid artery were catheterized with 12 F and 8 F ECMO catheters. The femoral artery and vein were cannulated for arterial blood pressure and central venous pressure monitorization, and for extraction of blood samples. Finally, the technique was used in six patients requiring ECMO, after obtaining informed consent from the parents. Anticoagulation was performed with heparin, always administering 5 U/kg/h into the hemofiltration circuit and the remainder (mean of 15.3 U/kg/h) into the ECMO circuit to maintain an activated coagulation time (Hemochron Jr II, ITC, Edison, USA), between 180 and 220 s. We thank Lucía González, Blanca Ramírez, and Maria Angeles Sierra, perfusionist nurses of the Cardiac Surgery Service, for helping us to perform the extracorporeal membrane oxygenation and continuous renal replacement therapy.