Ascites After Liver Transplantation
←
→
Page content transcription
If your browser does not render page correctly, please read the page content below
RAPID COMMUNICATION Ascites After Liver Transplantation Isabel Cirera,* Miguel Navasa,* Antoni Rimola,* Juan Carlos Garcı́a-Pagán,* Luis Grande,† Juan Carlos Garcia-Valdecasas,† Josep Fuster,† Jaime Bosch,* and Juan Rodes* Massive ascites after liver transplantation, although un- been reported to cause significant obstruction to graft common, usually represents a serious adverse event. The venous drainage and massive ascites formation after pathogenesis of this complication has not been ad- liver transplantation in some cases, although evident equately investigated. To determine the incidence, charac- teristics, and pathogenic factors of massive ascites after mechanical obstructions have not been found in the liver transplantation (ascitic fluid G 500 mL/d for G10 majority of patients with posttransplantation ascites.3,4 days), the charts of 378 liver transplant recipients were Decreased liver vascular compliance during acute cellu- reviewed. Massive ascites occurred in 25 patients (7%). lar rejection5,6 or the use of reduced grafts7 causing Mean ascitic fluid production was 960 mL/d (range, 625 inadequate accommodation of liver blood flow have to 2,350 mL/d), and the duration of ascites was 77 days also been proposed as mechanisms of ascites formation (range, 15 to 223 days). The ascitic fluid had a high protein content (36 6 7 g/L; range, 25 to 50 g/L). When in liver transplant recipients. Recently, Gane et al8 patients who did and did not develop massive ascites were reported a series of liver transplant recipients who compared, significant differences were found in receptor developed massive ascites and coagulation disturbances sex (men, 88% v 60%, respectively; P F .01) and surgical associated with an underlying hypercoagulable state, technique (inferior vena cava preservation with piggyback but no information on the outcome of these patients technique, 72% v 41%; P F .01). Significantly increased wedged and free hepatic venous pressures and gradients was given. However, despite these reports, the possible between hepatic vein and right atrial pressures were found factors involved in the pathogenesis of massive ascites in patients who developed ascites, suggesting a difficulty after liver transplantation have not been fully investi- in graft blood outflow. Massive ascites was associated with gated. renal impairment, increased incidence of abdominal infec- The current study investigates the incidence and tion, prolonged hospitalization, and a tendency toward characteristics of massive ascites after liver transplanta- reduced survival. In conclusion, massive ascites after liver transplantation is relatively uncommon but associated tion, as well as its impact on the outcome of liver with increased morbidity and mortality and is predomi- transplant recipients. In addition, systemic and splanch- nantly related to difficulties of hepatic venous drainage. nic hemodynamics and other donor and recipient Measurement of hepatic vein and atrial pressures to detect variables were compared between patients who did and a significant gradient and correct possible alterations in did not develop massive ascites to investigate the hepatic vein outflow should be the first approach in the possible mechanisms associated with the development management of these patients. Copyright r 2000 by the American Association for the of this complication. Study of Liver Diseases Patients and Methods O rthotopic liver transplantation is considered the best therapeutic option for end-stage liver dis- ease, with excellent survival and substantial improve- The records of 378 consecutive liver transplantations per- formed in 349 adult patients who survived more than 10 days after surgery were reviewed. Massive ascites was arbi- ment in quality of life.1 The postoperative period is trarily defined as the production of ascitic fluid greater than characterized by a high incidence of medical and 500 mL/d, assessed by loss of ascitic fluid through drain surgical complications, including graft rejection, bacte- tubes, surgical wounds, or paracentesis, that lasted longer rial infections, and vascular or biliary problems, which than 10 days after the surgical procedure. have been extensively described. However, the develop- ment of ascites after transplantation has received little From the *Liver Unit and †Digestive Surgery, Institut de Malalties attention. Small to moderate amounts of ascitic fluid Digestives, Hospital Clı́nic, Institut d’Investigacions Biomèdiques August are often observed in the early postoperative period but Pi i Sunyer (IDIBAPS), University of Barcelona, Spain. usually disappear in a few days.2 Nevertheless, large Address reprint requests to Miguel Navasa, MD, Liver Unit, Hospital Clı́nic, Villarroel 170, 08036 Barcelona, Spain. volumes and a long duration of ascites develop in some Copyright r 2000 by the American Association for the Study of patients. Liver Diseases Stenosis of the inferior caval vein anastomosis has 1527-6465/00/0602-0013$3.00/0 Liver Transplantation, Vol 6, No 2 (March), 2000: pp 157-162 157
158 Cirera et al To identify the possible factors related to the development The total ascitic fluid loss was 30 6 19 L (range, 10 of ascites, the following variables were analyzed: donor age; to 110 L), and the duration of ascites was 77 6 53 days pretransplantation recipient variables, including age, sex, (range, 15 to 223 days), with an average daily loss of liver disease, history of ascites, spontaneous bacterial peritoni- 960 mL per patient (range, 625 to 2,350 mL). At the tis or previous abdominal surgery, and laboratory data, time of detection of ascites, total ascitic fluid protein including standard renal and liver function test results; concentration was usually high (36 6 7 g/L; range, 25 perioperative data, including cold ischemia time, red blood cell transfusion requirements, type of inferior caval vein to 50 g/L), white and red blood cell counts in ascitic anastomosis (replacement of the recipient inferior vena cava fluid were variable (1,280 6 2,200 and 42,000 6 [classic technique] or preservation of inferior vena cava 79,000 cells/mL, respectively; range, 210 to 10,600 [piggyback technique]); peak serum alanine transferase value; and 180 to 290,000 cells/mL, respectively), depending and necessity of reoperation during the early postoperative on the postoperative day in which samples were period (first 3 days after surgery). collected, and ascitic fluid cultures obtained by sterile Hepatic and systemic hemodynamic data, obtained be- paracentesis and bedside inoculation of the fluid in tween the second and third week after liver transplantation, aerobic and anaerobic blood culture bottles showed were available in a subgroup of 38 patients. Hemodynamic negative results for all patients. studies were performed because patients either developed As listed in Table 1, patients who developed ascites massive ascites, were included on a previously published after transplantation were similar to those who did not study investigating hemodynamic changes after liver transplan- tation,9 or required a transjugular liver biopsy for the in relation to age, type of pretransplantation liver diagnosis of different graft alterations in the setting of severe disease, previous abdominal surgery, past history of coagulopathy that precluded a percutaneous liver biopsy. ascites or spontaneous bacterial peritonitis, and preop- Wedged and free hepatic venous pressures and right atrial erative renal and liver function test results. However, pressures were measured, and mean arterial pressure, cardiac male sex was more frequent in the ascitic than output, and systemic vascular resistance were obtained and nonascitic group (88% v 60%, respectively; P , .01). calculated according to previously described methods.9 Inten- As listed in Table 2, there were no significant differ- tionally, no large amount of ascitic fluid was present at the time of the hemodynamic studies to avoid the influence of increased abdominal cavity pressure on hemodynamic param- eters. For this purpose, in patients in whom spontaneous Table 1. Pretransplantation Demographic, Clinical, and removal of ascites through surgical wound or drains did not Laboratory Data From 378 Liver Transplant Recipients occur, large-volume paracentesis was performed before the Classified According to Development of Ascites After hemodynamic study. An intravenous albumin infusion (10 Transplantation g/L of ascites removed) was administered after each paracen- tesis. In all patients, the permeability of the portal and Ascites No Ascites inferior caval veins and hepatic artery was assessed by (n 5 25) (n 5 353) P repeated pulsed Doppler studies. Age (y) 47 6 11 46 6 11 NS Sex ,.01 Statistical Analysis Men 22 (88) 212 (60) Comparison of quantitative variables was made with Stu- Women 3 (12) 141 (40) dent’s t-test. Qualitative variables were compared by Chi- Liver disease NS squared test. Results are reported as mean 6 SD. Statistical Cirrhosis 20 (80) 266 (75) significance is established at P less than .05. Acute liver failure 2 (8) 37 (11) Retransplantation 2 (8) 36 (10) Other 1 (4) 14 (4) Results Previous abdominal surgery 9 (36) 167 (47) NS Massive ascites developed after surgery in 25 liver Previous ascites 18 (72) 212 (60) NS transplant recipients (7%). This group included 22 Previous SBP 5 (20) 55 (16) NS men and 3 women, with a mean age of 47 years. The Serum bilirubin (mg/dL) 8.8 6 11.1 5.7 6 6.0 NS indications for transplantation were cirrhosis in 20 Prothrombin activity (%) 63 6 26 57 6 23 NS Serum albumin (g/L) 30 6 7 32 6 6 NS patients (posthepatitic in 12 patients, alcoholic in 5 Serum creatinine (mg/dL) 0.9 6 0.3 0.9 6 0.7 NS patients, cryptogenic in 2 patients, and primary biliary in 1 patient) and miscellaneous liver diseases in the NOTE. Values expressed as mean 6 SD or number (per- cent). remaining 5 patients (acute liver failure in 2 patients, Abbreviations: SBP, spontaneous bacterial peritonitis; NS, retransplantation in 2 patients, and primary type I not significant. hyperoxaluria in 1 patient).
Ascites After Liver Transplantation 159 eters together, a gradient between free hepatic vein and Table 2. Perioperative and Donor Variables From 378 right atrial pressures greater than 6 mm Hg with a Liver Transplant Recipients Classified According to wedged hepatic vein pressure greater than 12 mm Hg Development of Ascites After Transplantation were observed in 100% of the patients with ascites and 18% of the patients without ascites (P , .001). Ascites No Ascites (n 5 25) (n 5 353) P All patients were initially treated with diuretics (spironolactone and/or furosemide) and intravenous Donor age (y) 39 6 18 33 6 15 NS albumin infusions. Furthermore, in 21 patients, re- Donor to recipient body weight ratio 1.01 6 0.20 1.09 6 0.20 NS peated large-volume paracenteses with intravenous Cold ischemia time (min) 410 6 193 384 6 177 NS albumin infusions were performed. These treatments Type of inferior vena cava did not have an apparent effect on the rate of ascitic anastomosis ,.01 fluid formation. In 1 patient, a peritoneovenous shunt Piggyback technique 18 (72) 145 (41) was inserted on day 32 after liver transplantation, with Classic technique 7 (28) 208 (59) rapid clinical improvement and resolution of ascites. RBC transfusion (units) 14 6 17 11 6 12 NS However, 1 month later, ascites reappeared and throm- Early reoperation 9 (36) 69 (19) NS bosis of the shunt was confirmed by radiological Early postoperative peak evaluation. Two days later, the piggyback anastomosis serum ALT (IU/L) 507 6 674 409 6 614 NS was reconstructed by performing a side-to-side cavoca- NOTE. Values expressed as mean 6 SD or number (per- val anastomosis. In the other 2 patients, side-to-side cent). Abbreviations: RBC, red blood cells; ALT, alanine aminotrans- cavocaval anastomoses were also performed. Ascites ferase; NS, not significant. rapidly resolved after the procedure in these 3 patients. In another patient with a stricture in the inferior vena cava anastomosis, a percutaneous transluminal angio- ences between the 2 groups of patients regarding cold plasty was successfully performed. In the remaining 15 ischemia time, perioperative red blood cell transfusion, patients with ascites who survived, ascites formation reoperation, and peak serum alanine aminotransferase slowly decreased and ascitic fluid finally disappeared values within the first 3 postoperative days. The without specific treatment. piggyback technique was performed more frequently All patients with ascites developed renal impair- in patients who developed ascites than in those who did not (72% v 41%, respectively; P , .01). Hemodynamic measurements were performed in Table 3. Hemodynamic Data of Patients Who Did and 11 patients with ascites and 27 patients without ascites. Did Not Develop Ascites After Liver Transplantation As listed in Table 3, there were no significant differ- ences in cardiac output, mean arterial pressure, sys- Ascites No Ascites temic vascular resistance, right atrial pressure, and (n 5 11) (n 5 27) P hepatic venous pressure gradient between the 2 groups. Mean arterial pressure Conversely, significantly greater free and wedged he- (mm Hg) 97 6 14 99 6 12 NS patic venous pressures and gradients between wedged Cardiac output (L/min) 5.6 6 1.5 6.2 6 1.2 NS hepatic vein and right atrial pressures and between free SVR (dyn · sec · hepatic vein and right atrial pressures were found in cm25) 1353 6 294 1436 6 436 NS patients who developed ascites, suggesting a difficulty RAP (mm Hg) 3.3 6 2.4 2.7 6 3.5 NS in graft blood outflow. As shown in Figure 1, all WHVP (mm Hg) 17.5 6 2.4 10.8 6 4.1 ,.0001 patients with ascites had a gradient between free FHVP (mm Hg) 13.2 6 3.3 8 6 4.6 ,.001 hepatic vein and right atrial pressures greater than 6 HVPG (mm Hg) 3.9 6 2.7 2.7 6 1.7 NS mm Hg, whereas this occurred in only 33% of the WHVP-RAP gradient patients without ascites (P , .01). In addition, wedged (mm Hg) 14.3 6 3.3 8.2 6 4.5 ,.001 hepatic venous pressure was significantly greater in FHVP-RAP gradient (mm Hg) 10.4 6 3.7 5.6 6 4.6 ,.01 patients who developed ascites than in those without ascites. All patients with ascites had a wedged hepatic Abbreviations: SVR, systemic vascular resistance; WHVP, wedged hepatic venous pressure; FHVP, free hepatic venous venous pressure greater than 12 mm Hg, but this pressure; HVPG, hepatic venous pressure gradient; RAP, occurred in only 22% of the patients without ascites right atrial pressure. (P , .001). Considering these 2 hemodynamic param-
160 Cirera et al Figure 1. (Left) Individual wedged hepatic venous pres- sure and (right) free hepatic venous to right atrial pressure gradient in 11 patients who developed ascites and 27 pa- tients who did not. ment during hospitalization. Serum creatinine levels patients who developed ascites compared with patients days 15 and 30 after transplantation were significantly who did not develop this complication (24% v 12%, (P , .001) greater in patients with ascites than in those respectively), although this difference did not reach without ascites: 2.6 6 1.4 versus 1.2 6 1.0 mg/dL and statistical significance (P 5 .16). Causes of death in 2.0 6 1.0 versus 1.1 6 0.6 mg/dL, respectively. Renal patients who developed ascites were sepsis with multi- impairment paralleled the course of ascites, with organ failure in all patients while ascites was still normalization of serum creatinine levels in patients in present. No differences were observed between patients whom ascites resolved. One year after liver transplanta- who died and survivors with regard to hemodynamic tion, serum creatinine levels were similar in survivors data. from both groups (1.2 6 0.2 v 1.3 6 0.4 mg/dL, respectively). Development of ascites was associated with a greater Discussion rate of peritonitis. The incidence of peritonitis was 8 of The current study reports an incidence of ascites after 25 patients with ascites (32%) and 19 of 348 patients liver transplantation of 7%. Trauma resulting from without ascites (6%; P , .001). No apparent intra- surgical manipulation has been involved in the patho- abdominal source of infection was identified in any genesis of this syndrome.2,8 However, this could be the patient with ascites and peritonitis, although in 5 explanation for ascites in the early postoperative pe- patients, peritonitis occurred while there was spontane- riod, which usually disappears in a few days, but this ous discharge of ascitic fluid through the surgical mechanism hardly applies to patients with massive and wound or drain sites. Conversely, peritonitis in patients long-lasting ascitic fluid formation. without ascites was related to biliary leakage (10 From the hemodymanic data obtained in our patients) and abdominal abscess (9 patients). The patients, a major mechanism for massive ascites forma- isolated bacteria from the ascitic fluid of patients with tion after transplantation appears to be postsinusoidal ascites and peritonitis were predominantly of cutane- portal hypertension secondary to hepatic vein outflow ous origin (5 patients with Staphylococcus epidermidis; 2 difficulty. This suggestion is supported by the finding patients, Staphylococcus aureus; and 1 patient, Streptococ- that the gradient between free hepatic vein and right cus faecalis), whereas in peritonitic patients without atrial pressures was significantly greater in patients who ascites, the causative organisms were predominantly of developed ascites than in patients who did not. It is enteric origin (13 patients with S faecalis; 2 patients, important to note that although the main problem of Escherichia coli; 2 patients, Pseudomonas spp; 1 patient, these patients is related to outflow difficulty, shown by Bacteroides fragilis; and 1 patient, S aureus). the increased free hepatic vein to right atrial pressure Ascites significantly prolonged hospital stay (59 6 gradient, massive ascites only appeared when the 45 days in patients with ascites and 44 6 26 days in wedged hepatic venous pressure, which mirrors sinusoi- patients without ascites; P , .01). There was a trend dal pressure,10 overcomes the threshold value of 12 mm toward a greater 1-year mortality rate in the group of Hg. Thus, the hemodynamic outflow difficulty only
Ascites After Liver Transplantation 161 becomes clinically relevant when it is important enough analogue) was also successful in 1 case of chylous to promote a backward increase in sinusoidal pressure. ascites after liver transplantation.16 Increased sinusoidal pressure would enhance the filtra- In the present study, ascites after liver transplanta- tion of fluid to the interstitium. Because the drainage tion was associated with marked renal impairment. In capacity of hepatic lymphatics is probably reduced or patients with important ascites production, fluid imbal- nonexistent because of the surgical dissection and ance caused by massive ascitic fluid loss, aggravated by ligation of the graft lymphatic vessels, interstitial fluid diuretic therapy in some occasions, could explain this would accumulate in the liver and finally in the renal impairment. Conversely, immunosuppressive peritoneal cavity. The high protein content in the therapy with cyclosporine or tacrolimus, as well as the ascitic fluid of these patients is in keeping with this eventual administration of other potentially nephro- hypothesis. Because a clear stenosis or thrombosis toxic drugs, could have contributed to renal impair- could be detected in only 1 of our patients with ascites, ment in these patients. Removal of drains and control it is likely that the hepatic venous outflow disturbance of wound leakage has been recommended to avoid was mostly related to a kinking of inferior caval vein or ascitic fluid loss.2 Replacement of fluid loss and graft malposition, causing insufficient venous drain- administration of albumin to maintain oncotic pres- age,10 particularly in our patients in whom the piggy- sure is also recommended. Nonetheless, it should be back technique was used. It should be noted that when noted that treatment aimed at correcting the problem caval anastomosis was reconstructed in some of our of hepatic vein outflow was followed by normalization of renal function in our patients. patients, ascites resolved promptly thereafter. Further- Another adverse consequence of ascites was an more, after the analysis of these results, we changed the increased risk for peritoneal infection, which was piggyback technique, which formerly only included 5-fold greater in patients who developed massive the origin of the left and medium hepatic veins and ascites than in patients who did not develop this now includes the origin of all 3 hepatic veins to complication. The existence of spontaneous discharge facilitate good graft venous drainage. After this techni- of ascitic fluid through surgical and drain wounds at cal modification, massive ascitic fluid loss in our the time of peritoneal infection in the majority of patients has become anecdotal. In this setting, the patients with massive ascites and the probable cutane- transjugular insertion of a metal stent into the hepatic ous origin of most organisms responsible for peritonitis venous outflow tract has also been reported as a in these patients suggest that the main mechanism of successful method to overcome the difficulty of the infection was the contamination of ascitic fluid from outflow drainage in liver transplant recipients who the skin. developed gross ascites after the surgery.4,11 Because massive ascites after transplantation is A contributory mechanism to ascites formation in associated with important morbidity, it is not surpris- these patients could be the increase in hepatic blood ing that patients who developed massive ascites had a flow commonly observed after liver transplantation. more prolonged hospital stay than patients who did This is caused by persistence of the splanchnic hyper- not develop this complication. Furthermore, patients emia associated with portal hypertension, the hepatic developing ascites showed a clear tendency toward a arterial vasodilatation caused by hepatic artery denerva- poorer survival, although not reaching statistical signifi- tion of the graft, or both mechanisms.9,12-14 In most cance, probably because of the small number of cases, the graft could accommodate this increased patients with massive ascites. blood flow. In other cases, however, the increase in In summary, although massive ascites is uncommon hepatic blood flow would increase the sinusoidal after liver transplantation, this complication is a severe pressure if a difficulty in hepatic venous outflow was event because it is associated with an increased risk for present and thereby contribute to the ascites forma- renal failure and peritoneal infection, prolonged hospi- tion.15 Whether this mechanism could be involved in tal stay, and possibly shortened survival. Data from the the development of ascites after liver transplantation present study suggest that ascites after liver transplanta- cannot be ascertained by this study, but if this were tion is predominantly related to a difficulty in hepatic true, the administration of drugs that reduce the venous outflow. Therefore, the measurement of hepatic splanchnic blood flow, such as somatostatin, proprano- vein and atrial pressures to detect significant pressure lol, or glypressin, may be useful in the therapy of this gradients and the adequate corrections of the alter- complication. Propranolol was effective in 1 case ations in hepatic vein outflow appear to be the first reported by Gane et al,8 and octreotide (a somatostatin approach in the management of these patients.
162 Cirera et al References 10. Bañares R, Castellote JI, Cuesta J, Casado M, Alvarez R, Cos E, et al. Ascites after orthotopic liver transplantation. A possible 1. United Network for Organ Sharing. UNOS Update. Rich- influence of piggyback anastomosis [abstract]. J Hepatol 1997; mond, VA: Scientific Registry, 1995. 26(suppl):A293S. 2. Howard TK. Postoperative intensive care management of the 11. Nolte W, Canelo R, Figulla HR, Kersten J, Sattler B, Munke H, adult. In: Busuttil RW, Klintmalm GB (eds). Transplantation of et al. Transjugular intrahepatic portosystemic stent-shunt after the liver (ed 1). Philadelphia: Saunders, 1996:551-563. orthotopic liver transplantation in a patient with early recur- 3. Zajko AB, Claus D, Clapuyt P, Esquivel CO, Moulin D, Starzl rence of portal hypertension of unknown origin. Z Gastroen- TE, et al. Obstruction to hepatic venous drainage after liver terol 1998;36:159-164. transplantation: Treatment with balloon angioplasty. Radiology 1989;170:763-765. 12. Henderson JM, Mackay GJ, Hooks M, Chezmar JL, Galloway 4. John TG, Plevris JN, Rehead DN, Garden OJ, Finlayson NDC, JR, Dodson TF, et al. High cardiac output of advanced liver Sanfey HA. Massive ascitic fluid loss after liver transplantation disease persists after orthotopic liver transplantation. Hepatol- (letter). Gastroenterology 1996;111:564-568. ogy 1992;15:258-262. 5. Noble-Jamieson G, Jamieson N, Barnes ND. Ultrafiltration for 13. Hadengue A, Lebrec D, Moreau R, Sogni P, Durand F, Gaudin intractable ascites after liver transplantation. Arch Dis Child C, et al. Persistence of systemic and splanchnic hyperkinetic 1991;66:988-989. circulation in liver transplant patients. Hepatology 1993;17:175- 6. Mabrut JY, de la Roche E, Adham M, Ducerf C, Baulieux J. 178. Peritoneovenous diversion using the LeVeen shunt in the 14. Gadano A, Hadengue A, Widmann JJ, Vachiery F, Moreau R, treatment of refractory ascites after liver transplantation. Ann Yang S, et al. Hemodynamics after orthotopic liver transplanta- Chir 1998;52:612-617. tion: Study of associated factors and long term effects. Hepatol- 7. Adetiloye VA, John PR. Intervention for pleural effusions and ogy 1995;22:458-465. ascites following liver transplantation. Pediatr Radiol 1998;28: 15. Benoit JN, Womack WA, Hernández LA, Granger DN. ‘‘For- 539-543. 8. Gane E, Langley P, Williams R. Massive ascitic fluid loss and ward’’ and ‘‘backward’’ flow mechanisms of portal hypertension. coagulation disturbances after liver transplantation. Gastroenter- Relative contributions in the rat model of portal vein stenosis. ology 1995;109:1631-1638. Gastroenterology 1985;89:1092-1096. 9. Navasa M, Feu F, Garcı́a-Pagán JC, Jiménez W, Llach J, Rimola 16. Shapiro AMJ, Bain VG, Sigalet DL, Kneteman NM. Rapid A, et al. Hemodynamic and humoral changes after liver resolution of chylous ascites after liver transplantation using transplantation in patients with cirrhosis. Hepatology 1993;17: somatostatin analog and total parenteral nutrition. Transplanta- 355-360. tion 1996;61:1410-1411.
You can also read