Complications Associated With Percutaneous Placement of Venous Return Cannula for Venovenous Bypass in Adult Orthotopic Liver Transplantation

Transcription

1 LIVER TRANSPLANTATION 13: , 2007 ORIGINAL ARTICLE Complications Associated With Percutaneous Placement of Venous Return Cannula for Venovenous Bypass in Adult Orthotopic Liver Transplantation Tetsuro Sakai, 1 Raymond M. Planinsic, 1 * Ibetsam A. Hilmi, 1 and J. Wallis Marsh 2 1 Department of Anesthesiology, University of Pittsburgh Medical Center (UPMC) Presbyterian/Montefiore Hospital, University of Pittsburgh Medical Center, University of Pittsburgh, Pittsburgh, PA; and 2 Department of Surgery, Thomas E. Starzl Transplantation Institute, University of Pittsburgh, Pittsburgh, PA Percutaneous large bore cannula placement during orthotopic liver transplantation (OLT) for use in venovenous bypass (VVB) has been reported to be a rapid and simple technique. It is, however, a technique that carries its own risks. The aim of the study was to investigate the incidence of complications related to the placement of a percutaneous venous return cannula and subsequent VVB in OLT. A retrospective review of 360 consecutive adult OLT patients during a period of 18 months (January 1, 2003 to June 30, 2004) was performed. The percutaneous venous cannula (18 Fr) was placed by an attending transplant anesthesiologist. The cannulation was attempted in 326 patients (90.6%). No cannulation was attempted on the subclavian veins. Internal jugular venous cannula placement was attempted but aborted in 6 patients (1.8%) due to technical difficulties. In 320 patients who received an internal jugular venous cannula, 313 (97.8%) underwent OLT without complication. The remaining 7 patients (2.2%) had complications. The operation was delayed for 1 patient due to suspected hemomediastinum. The other 6 complications were related to VVB: air embolism (2 patients), low flow rate (2 patients), hypotension (1 patient), and atrial fibrillation (1 patient). Successful OLT was eventually carried out in all the 7 patients and no mortality associated with internal jugular venous cannula placement or VVB was noted. In conclusion, percutaneous placement of a large bore venous return cannula for VVB during adult OLT can be performed with acceptable risk using a flexible 18-Fr cannula via the right internal jugular vein (IJV) by experienced attending transplant anesthesiologists. Liver Transpl 13: , AASLD. Received August 1, 2006; accepted October 27, The need for a bypass circuit during orthotopic liver transplantation (OLT) was first indicated by Moore et al. 1 in The current venovenous bypass (VVB) system consists of a centrifugal pump and heparin bonded tubing, and was introduced in 1983 by Griffith et al. 2 The technique has been adopted in many transplantation centers and has improved hemodynamic stability, reduced blood loss, and prolonged a tolerable anhepatic period. Recently, however, there has been a trend toward more selective use of VVB due to concerns of increased morbidity and mortality associated with VVB and placement of the cannula. 3,4 Traditionally, a venous return cannula was inserted into the axillary vein with surgical cut-down, which carried a high incidence of complications, 4 such as seromas or lymphoceles, 5,6 wound infection, and nerve injuries. 7 A percutaneous technique, first introduced in 1994, 8 has been shown to reduce complications related to the surgical cut-down, 9,10 maintain adequate shunt flow and stable hemodynamics, 9,11 and has become an Abbreviations: OLT, orthotopic liver transplantation; VVB, venovenous bypass; IJV, internal jugular vein; TEE, transesophageal echocardiography. Supported by University of Pittsburgh, Pittsburgh, PA. Presented at the annual meeting of International Liver Transplantation Society, Milan, Italy, May 3-6, Address reprint requests to Raymond M. Planinsic, MD, Director of Hepatic Transplantation Anesthesiology, University of Pittsburgh Medical Center, UPMC Presbyterian, C207, 200 Lothrop Street, Pittsburgh, PA Telephone: ; FAX: ; planinsicrm@anes.upmc.edu DOI /lt Published online in Wiley InterScience ( American Association for the Study of Liver Diseases.

2 962 SAKAI ET AL. TABLE 1. Demographic and Clinical Characteristics for 360 Liver Transplants Age [median (range) yr] 53 (18-76) Male gender [n (%)] 224 (62%) Isolated primary OLT 321 Isolated redo OLT 36 Graft failure 28 Primary nonfunctioning 8 Combined primary OLT and 3 thoracic transplantations OHT* 1 DLT 1 OHT DLT 1 Abbreviations: OLT, orthotopic liver transplantation; OHT, orthotopic heart transplantation; DLT, double lung transplantation. *Primary diagnosis for the combined transplantation was amyloidosis. Primary diagnosis for the combined transplantation was cystic fibrosis Primary diagnosis for the combined transplantation was idiopathic pulmonary hypertension. Figure 1. An 18-Fr Fem-Flex Duraflo-treated femoral arterial cannula (Baxter, Irvine, CA) (D) with 3 dilators (A-C). established alternative method for cannula placement in VVB. Unfortunately, the percutaneous method carries its own risks. Several case reports have described complications following the percutaneous VVB technique, including life-threatening complications and even mortality. 12,13 Systematic reviews focusing on the incidence of complications with this percutaneous method has been very limited. 12 The aim of this study was to review the incidence of morbidity and mortality associated with percutaneous VVB in adult OLT. Complications related to vascular access and those associated with subsequent use of extracorporeal circuit, were each analyzed. PATIENTS AND METHODS Data Collection All patients included in this study underwent OLT at the Starzl Transplant Institute from January 1, 2003 to June 30, All clinical data were prospectively collected and stored in the hospital wide electrical file system. The anesthesia record, the perfusion record, and other in-hospital clinical records of each patient were retrospectively reviewed. Demographic features of the 360 patients who had an OLT during this 18-month period and included in this study are shown in Table 1. Veno-Venous Cannula Placement and Veno- Venous Bypass After the induction of the general anesthesia and endotracheal intubation, an 18-Fr Fem-Flex Duraflo Treated Femoral Arterial Cannula (18 Fr 15 cm; Baxter, Irvine, CA) (Fig. 1) was inserted percutaneously into the internal jugular vein (IJV) in the following fashion. The Figure 2. An 18-Fr Fem-Flex Duraflo-treated femoral arterial cannula was inserted on the right IJV. A tube clamp was being placed in the middle of the cannula after removal of the stylet and the guide wire. IJV was first located using a small needle (20 gauge), the aid of surface anatomy and with the patient placed in Trendelenberg position. Using the Seldinger technique, the entry to the IJV was dilated with 3 dilators of increasing size (Fig. 1) prior to insertion of the 18-Fr cannula (Fig. 2). Aspiration of the venous blood and infusion of normal saline through the cannula were used to confirm IJV placement. This procedure was performed by an attending transplant anesthesiologist (8 physicians total during the period). The cannula was secured with a silk stitch and continuously flushed with normal saline until the commencement of VVB. A transesophageal echocardiography (TEE) probe (Omniplane II; Hewlett Packard) was inserted in all cases after cannulation, and attempt was made to confirm the cannula position in the superior vena cava. When it was difficult to identify the tip of the cannula with this modality, a bubble test was performed (injecting the well-

3 PERCUTANEOUS VVB CANNULA COMPLICATIONS IN OLT 963 mixed solution of 9 ml of normal saline with 1 ml of air into the return cannula followed by normal saline flush, while monitoring the 4-chamber view of TEE). Immediate appearance of microbubble image in the right atrium confirmed the location of the tip of the cannula in the superior vena cava. A chest X-ray was not routinely performed to verify location of the inflow cannula. Simultaneously, an 8.5-Fr venous cannula (Trauma Kit; Arrow International, Reading, PA) was inserted percutaneously in the left femoral vein by the surgical team to secure the venous access of the systemic drainage route for VVB. Other routine vascular accesses were established prior to the operation: 2 arterial lines (the radial and the right femoral artery), a central line (9-Fr MAC, Two-Lumen Central Venous Access Kit, Arrow International) with a Swan-Ganz catheter (8-Fr CCOmbo V; Edwards Lifesciences, Irvine, CA), and an additional volume infusion line (a 7-Fr RIC catheter [Arrow International] at the antecubital vein or a 9-Fr catheter [MAC, Two-Lumen Central Venous Access Kit, Arrow International] at the left IJV). After skeletonization of the liver, the 8.5-Fr femoral venous cannula was exchanged to a Bio-Medicus Femoral Venous kit (17 Fr 50 cm straight tip cannula; Medtronic, Anaheim, CA) for systemic drainage for VVB. For the portal venous drainage, a 26 Fr 50 cm straight tip Bio-Medicus Venous Cannula (Medtronic) was directly inserted and secured into the stump of the portal vein. The heparin-coated bypass circuit was established with a model CB 4649 custom pack liver set (3/8-inch tubing; Medtronic) and a Biomedicus BP-80 centrifugal pump (Medtronic). No heat exchanger was used. The priming solution of the circuit was 20% albumin (440 ml) at room temperature. The VVB machine was managed by a designated perfusionist. At the initiation of the VVB, the attending transplant anesthesiologist monitored a 4-chamber view of the heart with TEE and ensured the immediate and transient visualization of a microbubble image in the right atrium, as well as the stable chamber size and wall motion to rule out any extravasation of the return blood into the thoracic cavity, the pericardial space, or mediastinum. Total VVB flow and partial VVB flow with temporary occlusion of the portal limb line were measured and the cannula position was adjusted to achieve the optimal flow. Volume status and cardiac function were frequently monitored with TEE thereafter. The donor liver implantation was conducted in the orthotopic position in the piggyback fashion. After completing the anastomosis between the suprahepatic superior vena cava of the donor liver and the hepatic vein cuff of the recipient, the portal limb of the VVB was clamped and the portal cannula was removed for portoportal venous anastomosis. Volume status, femoral venous cannula position, pulmonary vascular resistance, and the speed of the centrifugal pump were adjusted to maintain more than 1.0 L/minute of VVB flow during this stage. At the completion of the porto-portal venous anastomosis, reperfusion of the donor liver was initiated with releasing first the portal and then the hepatic venous clamps. At the completion of reperfusion of the liver, the VVB was terminated. After disconnection from the VVB circuit, the venous return cannula in the IJV was again used as a fluid infusion line. At the completion of the surgery, the venous return cannula was removed and the insertion site of the skin was closed with a single silk purse-string suture by the surgical team. Manual compression on the suture site for 4 to 5 minutes was applied to complete the hemostasis. RESULTS Percutaneous VVB cannula placement was attempted on 326 of 360 patients. Among the 34 patients who did not receive a percutaneous VVB cannula, an axillary cut-down technique was employed as the initial choice in 5 patients. Reasons for axillary placement of the VVB cannula included a preexisting hemodialysis catheter in the right IJV in 3 patients, intracranial hypertension in 1, and obesity in 1 (body mass index 41.3). No VVB was used in 28 patients, mainly due to surgeons preference. In this group, 5 patients who underwent retransplantation for primary nonfunctioning were included. One patient underwent combined OLT and orthotopic heart transplantation for amyloidosis, where the venous return cannula was directly inserted in the right atrium for VVB during OLT. All the VVB cannulation attempts were made on the IJVs. The percutaneous cannulation for VVB was aborted in 6 patients due to technical difficulties, including inability to pass a guide wire, kinking of a guide wire, or difficulty in advancing a dilator. OLT was performed using axillary VVB bypass in 2 of these 6 patients, and no VVB was used in remaining 4 patients. The other 320 patients received IJV cannula placement. The VVB cannulas were placed in the right IJV in 310 patients and the left IJV in 10. The mean standard deviation of the VVB flow was 3, ml/ minute (1,000-5,000 ml/minute). Seven patients (2.2%) developed complications related to VVB or cannula placement for VVB (Table 2). Complications in all but 1 patient were related to extracorporeal circulation during VVB, including air embolism in 2 patients, low flow status (defined as less than 70% of the target VVB flow, which was calculated as 40 ml/kg/minute) in 2, systemic hypotension (defined as the significant decrease of the systemic blood pressure which indicated in shock status and necessitated aggressive resuscitation with the termination of VVB) in 1, and new onset of atrial fibrillation in 1. Only 1 complication of suspected hemomediastinum was directly related to the placement of the percutaneous return cannula. In this case, the cannula was placed via the right IJV without difficulty. Immediately after the placement of the cannula, however, persistent hypotension and tachycardia were noted. A chest X-ray showed questionable widening of the mediastinum. The patient was promptly sent to have a chest computed tomography, which was negative for hemomediastinum. OLT was performed with VVB using an axillary venous return cannula. All the 7 patients had a successful OLT. No mortality

4 964 SAKAI ET AL. TABLE 2. Summary of 7 Patients With Complications Under Veno-Venous Bypass and Its Venous Return Cannula Placement in Orthotopic Liver Transplantation Case Gender Age (yr) Site Complication Consequence 1 Male 36 RIJ s/o hemomediastinum A chest CT was negative; VVB was reestablished via the axillary vein 2 Female 52 RIJ Air embolism VVB was stopped, air was purged from VVB, then VVB was resumed 3 Male 52 RIJ Air embolism Bubbles in the right atrium on TEE; VVB was continued without any problem 4 Female 38 RIJ Low flow (1,000-1,500 ml/ minute) The target flow was 2,680 ml/minute; VVB was continued until reperfusion 5 Female 64 RIJ Low flow (1,400 ml/minute) The target flow was 2,100 ml/minute; VVB was terminated in 5 minutes; OLT was performed without VVB 6 Female 49 RIJ Hypotension (40/15 mmhg) VVB was terminated in 2 minutes; OLT was performed without VVB 7 Male 38 RIJ Atrial fibrillation VVB was continued; antiarrhythmic therapy was initiated; NSR was restored. Abbreviations: s/o, suspected of; RIJ, right internal jugular vein; CT, computed tomography; VVB, veno-venous bypass; TEE, transesophageal echocardiography; OLT, orthotopic liver transplantation; NSR, normal sinus rhythm. Figure 3. Disposition of the patients who underwent orthotopic liver transplantation during 18-month study period (January 1, 2002 to June 30, 2004). Abbreviations: OLT, orthotopic liver transplantation; PCVVB, percutaneous venovenous bypass; IJVC, internal jugular vein cannulation. was noted directly due to the cannula placement or VVB in this series. DISCUSSION During an 18-month period, of the 360 patients presenting for OLT, 320 patients had a percutaneous VVB return cannula placed via the IJV. The morbidity associated with placement of the percutaneous venous return cannula was 1 out of 320 (0.3%) and subsequent VVB related morbidity was 6 out of 320 (1.9%) (Fig. 3). No mortality was noted related to percutaneous cannula placement or successive VVB use. Studies reporting complications related to the percutaneous VVB return cannula placement have been limited in the past. In 2001, Budd et al. 12 reported that in 312 patients over an 8-yr period where a percutaneous technique was used for large bore cannula (18-20 Fr) placement, 4 patients suffered serious morbidity with hemothorax (1.28%). In this study, 1 death (0.32%) was directly related to the placement of percutaneous cannula. The cause was massive bleeding from vascular injury at the junction of the left subclavian vein and the left IJV after a technically difficult attempt of cannula insertion. 12 Recently, a fatal case was reported describing difficulty in placement of a 21-Fr cannula via the right IJV, which was followed by cardiac arrest upon commencement of VVB. Postmortem examination revealed the bypass catheter tip in the pleural space and a large right hemothorax. 13 There were suggestions of unpublished reports of deaths directly resulting from vascular trauma following percutaneous central venous cannulation, 14 but these could not be confirmed. A review of the cases with significant morbidity and mortality reveals some anecdotal information which may predispose patients to subsequent complications: use of a large bore cannula (more than 20-Fr reference), attempted access via a subclavian vein, 12 and technically difficult cannula insertions. 12,13 However, there has been no comparative study to prove these suggestions. During the period of this study, a policy was implemented regarding the percutaneous return cannula placement for VVB: 1) to be exclusively performed by the attending transplant anesthesiologists; 2) to use the right IJV as the preferred entry vein; 3) to use a smaller bore flexible cannula (18 Fr); and 4) to have low threshold to abort the percutaneous insertion of the bypass cannula if one encounters any technical difficulty. It has been suggested that the use of TEE guidance to facilitate the placement of the cannula would reduce complications. 12,15 However, this procedure was not

5 PERCUTANEOUS VVB CANNULA COMPLICATIONS IN OLT 965 used routinely during this study period, although the insertion of a TEE probe and monitoring the cardiac function throughout the liver transplantation has been a routine practice in our institution. A tube transducer technique was also used to confirm venous cannulation and avoid arterial cannulation. 16 This technique involves attaching a 36-inch monitor tubing (ARROW, Product #W01233; Arrow International) to an 18-G angiocatheter inserted into the IJV. If the angiocatheter is not kinked at the skin, then blood will easily flow into the monitor tubing when it is lowered. In addition, blood within the vertically held monitor tubing will move with respiration. These observations will confirm that the 18-G angiocatheter is indeed in the IJV. Using the above policy and the tube transducer technique, no inadvertent entry into the arterial vascular system with the 18-Fr cannula or its dilators was noted in this series. Guidance using real time 2-dimensional ultrasonography was not routinely used during the study period. Complications related to the VVB use after percutaneous cannula placement were also included in this study. Cases of fatal pulmonary embolism have resulted from thrombus either forming in the extracorporeal circuit or being translocated from the IVC to the right atrium. 4,17 Blood clotting in the bypass system may occur if the flow rate is very low ( 1 L/minute). 18 In this retrospective review, 6 patients developed complications during VVB, including air embolism, atrial fibrillation, and low flow states. All complications were managed as described in Table 2 without mortality. TEE has been reported to facilitate the diagnosis of the complications including the etiology of hemodynamic instability in patients undergoing noncardiac surgery. 19,20 All patients receiving OLT in this study were monitored continuously with TEE throughout the procedure. At the institution of the VVB, TEE was used to identify microbubbles entering the right atrium. In addition, stable cardiac chamber size and wall motion were confirmed at this stage. TEE was then used for monitoring venous status and cardiac function during VVB period, at reperfusion of the grafted liver and through the end of the procedure. In conclusion, percutaneous large bore return cannula placement for veno-venous bypass during adult OLT can be performed by experienced transplant anesthesiologists within acceptable risk using a flexible 18-Fr cannula via the right IJV. The anesthesiologists and surgical team are encouraged to have a low threshold to modify the technique for successful OLT if any technical difficulty is noted during the insertion of percutaneous venous return cannula placement via the IJV. REFERENCES 1. Moore FD, Wheele HB, Desmissianos HV, Smith LL, Balankura O, Abel K, et al. Experimental whole-organ transplantation of the liver and of the spleen. Ann Surg 1960; 152: Griffith BP, Shaw BW, Jr, Hardesty RL, Iwatsuki S, Bahnson HT, Starzl TE. Veno-venous bypass without systemic anticoagulation for transplantation of the human liver. Surg Gynecol Obstet 1985;160: Schumann R. Intraoperative resource utilization in anesthesia for liver transplantation in the United States: a survey. Anesth Analg 2003;97: Chari RS, Gan TJ, Robertson KM, Bass K, Camargo CA, Jr, Greig PD, et al. Venovenous bypass in adult orthotopic liver transplantation: routine or elective use? J Am Coll Surg 1998;186: Johnson SR, Marterre WF, Alonso MH, Hanto DW. A percutaneous technique for venovenous bypass in orthotopic cadaver liver transplantation and comparison with the open technique. Liver Transpl Surg 1996;2: Ozaki CF, Langnas AN, Bynon JS, Pillen TJ, Kangas J, Vogel JE, et al. A percutaneous method for venovenous bypass in liver transplantation. Transplantation 1994;57: Katirji MB. Brachial plexus injury following liver transplantation. Neurology 1989;39: Oken AC, Frank SM, Merritt WT, Fair J, Klein A, Burdick J, et al. A new percutaneous technique for establishing venous bypass access in orthotopic liver transplantation. J Cardiothorac Vasc Anesth 1994;8: Tisone G, Mercadante E, Dauri M, Colella D, Anselmo A, Romagnoli J, et al. Surgical versus percutaneous technique for veno-venous bypass during orthotopic liver transplantation: a prospective randomised study. Transplant Proc 1999;31: Mayoral V, Sabate A, Benito C, Camprubi I. Percutaneous femoro-porto-jugular venovenous shunt in orthotopic liver transplantation. Rev Esp Anestesiol Reanim 1996; 43: Washburn WK, Lewis WD, Jenkins RL. Percutaneous venovenous bypass in orthotopic liver transplantation. Liver Transpl Surg 1995;1: Budd JM, Isaac JL, Bennet J, Freeman JW. Morbidity and mortality associated with large-bore percutaneous venovenous bypass cannulation for 312 orthotopic liver transplantations. Liver Transpl 2001;7: Jankovic Z, Boon A, Prasad R. Fatal haemothorax following large-bore percutaneous cannulation before liver transplantation. Br J Anaesth 2005;95: Reddy K, Mallett S, Peachey T. Venovenous bypass in orthotopic liver transplantation: time for a rethink? Liver Transpl 2005;11: Planinsic RM, Nicolau-Raducu R, Caldwell JC, Aggarwal S, Hilmi I. Transesophageal echocardiography-guided placement of internal jugular percutaneous venovenous bypass cannula in orthotopic liver transplantation. Anesth Analg 2003;97: Marymont JH, Szokol JW, Murphy GS. Arterial or venous cannulation? A simple yet reliable detection technique. J Cardiothorac Vasc Anesth 1998;12: Navalgund AA, Kang Y, Sarner JB, Jahr JS, Gieraerts R. Massive pulmonary thromboembolism during liver transplantation. Anesth Analg 1988;67: Shaw BW, Jr, Martin DJ, Marquez JM, Kang YG, Bugbee AC, Jr, Iwatsuki S, et al. Venous bypass in clinical liver transplantation. Ann Surg 1984;200: Suriani RJ, Neustein S, Shore-Lesserson L, Konstadt S. Intraoperative transesophageal echocardiography during noncardiac surgery. J Cardiothorac Vasc Anesth 1998;12: Hofer CK, Zollinger A, Rak M, Matter-Ensner S, Klaghofer R, Pasch T, et al. Therapeutic impact of intra-operative transoesophageal echocardiography during noncardiac surgery. Anaesthesia 2004;59:3-9.