Pediatr Transplantation 2015: 19: 595–604
© 2015 John Wiley & Sons A/S. Published by John Wiley & Sons Ltd
Pediatric Transplantation DOI: 10.1111/petr.12549
Examinations of ascites from prophylactic drains can predict intra-abdominal infections after living donor liver transplantation Sanada Y, Kawano Y, Urahashi T, Ihara Y, Wakiya T, Okada N, Yamada N, Hirata Y, Tashiro M, Mizuta K. (2015) Examinations of ascites from prophylactic drains can predict intra-abdominal infections after living donor liver transplantation. Pediatr Transplant, 19: 595–604. DOI: 10.1111/petr.12549.
Yukihiro Sanada1, Youichi Kawano2, Taizen Urahashi1, Yoshiyuki Ihara1, Taiichi Wakiya1, Noriki Okada1, Naoya Yamada1, Yuta Hirata1, Masahisa Tashiro1 and Koichi Mizuta1 1
Abstract: Studies suggest that prophylactic intra-abdominal drains are unnecessary for cadaveric liver transplantation using whole liver grafts because there is no benefit from drainage. However, no studies have investigated on the necessity of prophylactic drains after LDLT using split-liver grafts or reduced-liver grafts, which may present a high risk of post-transplant intra-abdominal infections. This retrospective study investigated whether the ascitic data on POD 5 after LDLT can predict intra-abdominal infections and on the post-transplant management of prophylactic drains. Between March 2008 and March 2013, 90 LDLTs were performed. We assessed the number of ascitic cells, biochemical examinations, and cultivation tests at POD1 and POD5. The incidence rates of post-transplant intra-abdominal infections were 24.4%. The multivariate analysis showed that left lobe and S2 monosegment grafts were a significant risk factor for intra-abdominal infections (p = 0.006). The patients with intra-abdominal infections had significantly higher acsitic LDH levels and the positive rate of ascitic culture at POD5 in comparison with patients without infections (p < 0.001 and p = 0.014, respectively). LDLT using left lobe and S2 monosegment grafts yields a high risk for post-transplant intra-abdominal infections, and ascitic LDH and cultivation tests at POD5 via prophylactic drains can predict intra-abdominal infections.
LT is an established curative treatment for pediatric patients with end-stage liver disease or acute liver failure (1–3). However, despite improvements in surgical techniques and post-transplant care, post-transplant infection is one of the most common complications and a frequent cause of death in liver transplant patients (4, 5). An important category of infection at an early stage after LT is intra-abdominal infection (4–7). Intra-abdominal infections often occur at the surgical site and may be associated with prolonged
Abbreviations: AUC, area under the curve; CI, confidence interval; LDH, lactate dehydrogenase; LDLT, living donor liver transplantation; LT, liver transplantation; MP, methylprednisolone; Na, sodium; POD, post-operative day; ROC, receiver operating characteristic; Tac, tacrolimus.
Department of Transplant Surgery, Jichi Medical University, Shimotsuke City, Japan, 2Department of Surgery, Nippon Medical School, Bunkyo-ku, Japan
Key words: intra-abdominal infection – prophylactic intra-abdominal drain – ascitic examination – living donor liver transplantation Yukihiro Sanada, Department of Transplant Surgery, Jichi Medical University, 3311-1 Yakushiji, Shimotsuke City, Tochigi 329-0498, Japan Tel./Fax: +81 285 58 7069 E-mail: [email protected] Accepted for publication 30 May 2015
or complicated surgeries, or may be due to secondary infection of persistent ascites after LT (7, 8). Peritonitis or massive ascites at an early stage after LT can cause vascular complications (8–10) and may therefore be considered a poor prognostic factor (11, 12). Traditionally, in patients who undergo hepatic resection, intra-abdominal drains are routinely inserted into the subphrenic or subhepatic space near the transection site of liver. This procedure releases the intra-abdominal tension due to ascitic fluid accumulation and facilitates monitoring the occurrence of post-operative intra-abdominal bleeding, as well as the detection and drainage of any bile leakage (8). The risk of post-operative bleeding may also increase if there are associated portal hypertension, thrombocytopenia, and impaired 595
Sanada et al.
clotting profiles in patients with liver cirrhosis. The risk of post-operative bile leakage may also increase if there is associated LT using splitliver grafts or reduced-liver grafts. Moreover, the risk of post-operative ascites may increase if there are associated portal hypertension, outflow blockage, and acute cellular rejection in LT patients. Therefore, during liver transplant surgery using split-liver grafts or reduced-liver grafts, prophylactic intra-abdominal drains are inserted as information drains, and they may occasionally play an important role as prophylactic drainage devices for bleeding, bile leakage, intra-abdominal infections, and massive ascites (13). However, accumulating evidence suggests that prophylactic intra-abdominal drains are unnecessary for both gastrointestinal and liver surgeries because these surgical procedures can be performed safely without drainage (14–17). Drains limit the patient’s activity of daily living scale due to pain at the insertion site and the necessity of a drainage bag; moreover, a longterm placement of the drains can lead to a high risk of intra-abdominal infections by providing a route for ascending infection. In addition, drains after hepatic resection are associated with a significantly longer hospitalization and higher post-operative morbidity in patients with chronic liver diseases (14). Similar results demonstrating a lack of benefit from prophylactic intra-abdominal drains have been reported in patients undergoing cadaveric LT using wholeliver grafts (18–21). However, no studies have investigated on the necessity of prophylactic drains after living donor liver transplantation (LDLT) using split-liver grafts and reduced-liver grafts. LT using split-liver grafts and reducedliver grafts yields a possibility of bleeding and bile leakage from the transection site of graft liver in comparison with whole-liver grafts, and therefore may present a high risk of post-transplant intra-abdominal infections. Therefore, prophylactic intra-abdominal drains and ascitic examinations may be important tools for early diagnosis and treatment of bleeding, bile leakage, and intra-abdominal infections after LDLT. Only our group has previously reported ascitic data after LDLT (13), but there have been no widely agreed-upon methods for the efficacy and management of prophylactic post-transplant intra-abdominal drains. Therefore, the present retrospective study investigated whether ascitic data at POD 5 after LDLT can predict intraabdominal infections and on the post-transplant management of prophylactic drains. 596
Patients and methods Patients Between March 2008 and March 2013, 90 LDLTs were performed on pediatric patients using grafts associated with left lobe at the Department of Transplant Surgery, Jichi Medical University, Japan. The baseline demographic data pertaining to the recipients and grafts are presented in Table 1. Seventeen patients were excluded because insufficient data were collected at POD5 in 12 patients and urgent laparotomy or endovascular treatment until POD5 was undergone due to intra-abdominal bleeding in three patients (POD1, POD2, and POD2), hepatic arterial complication in one (POD4), and abdominal closure in one (POD5). Therefore, all of the recipients exhibited no specific complications (including surgical treatment) until POD5 in this study.
Immunosuppressive therapy Tac and MP were used as the standard post-operative immunosuppressive therapy. The target trough level of Tac was 15–20 ng/mL during the first week, 8–12 ng/mL during the first month, 5–8 ng/mL during the first six months, 3– 5 ng/mL during the first year, and 2–4 ng/mL thereafter. MP was administered at an initial dose of 20 mg/kg intravenously on the morning of the operation and before graft reperfusion. The MP dose was thereafter decreased gradually to three mg/kg/day on POD1, 0.5 mg/kg/day on POD7, and 0.25 mg/kg/day at one month post-LDLT, and was discontinued within one year, except in patients for whom immunosuppression could not be maintained at the lower dose. Mycophenolate mofetil was used when more potent immunosuppression was required, for example, in ABO-incompatible recipients older than five yr, in patients with steroid-resistant acute rejection episodes, and in patients with liver dysfunction after the cessation of MP therapy. Table 1. Demographic data of recipients and grafts Recipient and graft characteristics Gender Age (months) Body weight (kg) Original disease
PELD or MELD ABO compatibility Graft type GV/SLV (%) Operation time Cold ischemic time Warm ischemic time Blood loss volume (mL/kg) Transfusion volume (mL/kg)
Male 41, female 49 17 (0–234) 9.5 (2.6–53.7) Biliary atresia 66, ornithine transcarbamylase deficiency 10, fulminant hepatic failure 3, congenital extrahepatic portosystemic shunt 3, Alagille syndrome 2, graft hepatic failure 2, primary sclerosing cholangitis 2, citrullinemia 1, methylmalonic acidemia 1 9.4 ( 10.9–32.9) Identical 54, compatible 21, incompatible 15 Left lateral segment 60, left lobe 21, S2 monosegment 2, left lobe+caudate lobe 4 71.0 (33.0–130.0) 12 h 30 min (seven h two min–25 h 56 min) one h 54 min (46 min–six h 45 min) 40 min (24 min–two h two min) 75.7 (1.6–589.1) 98.7 (6.3–635.3)
PELD, pediatric end-stage liver disease; MELD, model for end-stage liver disease; GV/SLV, ratio of graft volume to standard liver volume.
Sanada et al.
various recipient and graft variables on intraabdominal infections was assessed, and the results are summarized in Table 3. The univariate analysis revealed that left lobe and S2 monosegment grafts, graft volume to standard liver volume ratio, ≥12 h 30 min of operation time, and ≥two h of cold ischemic time (p < 0.001, p = 0.043, p = 0.026, and p = 0.058, respectively) were independent risk factors for intraabdominal infections. The multivariate analysis revealed that left lobe and S2 monosegment grafts were an independent risk factor for intraabdominal infections (p = 0.006, odds ratio 10.10, and 95% confidence interval [CI] 1.91– 53.80) (Table 4). Ascitic examinations at POD5 in the patients with posttransplant intra-abdominal infections
The patients with post-transplant intra-abdominal infections had significantly higher acsitic neutrophil count, total bilirubin, and LDH levels at POD5 in comparison with patients without infections (p = 0.018, p = 0.018, and p = 0.001, respectively) (Fig. 1a–c). Consequent to ROC analysis of ascitic neutrophil count, total bilirubin, and LDH levels at POD5 in post-transplant intra-abdominal infections, the recommended cutoff value for diagnosing intra-abdominal infections was set at 624 per mm3, 2.3 mg/dL, and 536 IU/L, respectively. The ROC analysis of the ascitic neutrophil count had a sensitivity of 63.6%, specificity of 77.9%, AUC of 0.668, and 95% CI of 0.528–0.813 (Fig. 2a). The ROC analysis of the ascitic total bilirubin level had a sensitivity of 81.0%, specificity of 51.5%, the AUC of 0.645, and 95% CI of 0.500–0.791 (Fig. 2b). The ROC analysis of the ascitic LDH level had a sensitivity of 61.9%, specificity of 87.9%, the AUC of 0.748, and 95% CI of 0.618–0.877 (Fig. 2c).
Table 4. Risk factors of intra-abdominal infections after living donor liver transplantation: multivariate analysis Variables
Odds ratio
95% CI
p-value
Graft type Others vs. left lateral segment graft GV/SLV
10.10
1.91–53.80
0.006
20.50
0.36–1170.00
0.143
2.38
0.54–10.50
0.251
3.09
0.82–11.7
0.097
Operation time ≥12 h 30 min vs.
© 2015 John Wiley & Sons A/S. Published by John Wiley & Sons Ltd
Pediatric Transplantation DOI: 10.1111/petr.12549
Examinations of ascites from prophylactic drains can predict intra-abdominal infections after living donor liver transplantation Sanada Y, Kawano Y, Urahashi T, Ihara Y, Wakiya T, Okada N, Yamada N, Hirata Y, Tashiro M, Mizuta K. (2015) Examinations of ascites from prophylactic drains can predict intra-abdominal infections after living donor liver transplantation. Pediatr Transplant, 19: 595–604. DOI: 10.1111/petr.12549.
Yukihiro Sanada1, Youichi Kawano2, Taizen Urahashi1, Yoshiyuki Ihara1, Taiichi Wakiya1, Noriki Okada1, Naoya Yamada1, Yuta Hirata1, Masahisa Tashiro1 and Koichi Mizuta1 1
Abstract: Studies suggest that prophylactic intra-abdominal drains are unnecessary for cadaveric liver transplantation using whole liver grafts because there is no benefit from drainage. However, no studies have investigated on the necessity of prophylactic drains after LDLT using split-liver grafts or reduced-liver grafts, which may present a high risk of post-transplant intra-abdominal infections. This retrospective study investigated whether the ascitic data on POD 5 after LDLT can predict intra-abdominal infections and on the post-transplant management of prophylactic drains. Between March 2008 and March 2013, 90 LDLTs were performed. We assessed the number of ascitic cells, biochemical examinations, and cultivation tests at POD1 and POD5. The incidence rates of post-transplant intra-abdominal infections were 24.4%. The multivariate analysis showed that left lobe and S2 monosegment grafts were a significant risk factor for intra-abdominal infections (p = 0.006). The patients with intra-abdominal infections had significantly higher acsitic LDH levels and the positive rate of ascitic culture at POD5 in comparison with patients without infections (p < 0.001 and p = 0.014, respectively). LDLT using left lobe and S2 monosegment grafts yields a high risk for post-transplant intra-abdominal infections, and ascitic LDH and cultivation tests at POD5 via prophylactic drains can predict intra-abdominal infections.
LT is an established curative treatment for pediatric patients with end-stage liver disease or acute liver failure (1–3). However, despite improvements in surgical techniques and post-transplant care, post-transplant infection is one of the most common complications and a frequent cause of death in liver transplant patients (4, 5). An important category of infection at an early stage after LT is intra-abdominal infection (4–7). Intra-abdominal infections often occur at the surgical site and may be associated with prolonged
Abbreviations: AUC, area under the curve; CI, confidence interval; LDH, lactate dehydrogenase; LDLT, living donor liver transplantation; LT, liver transplantation; MP, methylprednisolone; Na, sodium; POD, post-operative day; ROC, receiver operating characteristic; Tac, tacrolimus.
Department of Transplant Surgery, Jichi Medical University, Shimotsuke City, Japan, 2Department of Surgery, Nippon Medical School, Bunkyo-ku, Japan
Key words: intra-abdominal infection – prophylactic intra-abdominal drain – ascitic examination – living donor liver transplantation Yukihiro Sanada, Department of Transplant Surgery, Jichi Medical University, 3311-1 Yakushiji, Shimotsuke City, Tochigi 329-0498, Japan Tel./Fax: +81 285 58 7069 E-mail: [email protected] Accepted for publication 30 May 2015
or complicated surgeries, or may be due to secondary infection of persistent ascites after LT (7, 8). Peritonitis or massive ascites at an early stage after LT can cause vascular complications (8–10) and may therefore be considered a poor prognostic factor (11, 12). Traditionally, in patients who undergo hepatic resection, intra-abdominal drains are routinely inserted into the subphrenic or subhepatic space near the transection site of liver. This procedure releases the intra-abdominal tension due to ascitic fluid accumulation and facilitates monitoring the occurrence of post-operative intra-abdominal bleeding, as well as the detection and drainage of any bile leakage (8). The risk of post-operative bleeding may also increase if there are associated portal hypertension, thrombocytopenia, and impaired 595
Sanada et al.
clotting profiles in patients with liver cirrhosis. The risk of post-operative bile leakage may also increase if there is associated LT using splitliver grafts or reduced-liver grafts. Moreover, the risk of post-operative ascites may increase if there are associated portal hypertension, outflow blockage, and acute cellular rejection in LT patients. Therefore, during liver transplant surgery using split-liver grafts or reduced-liver grafts, prophylactic intra-abdominal drains are inserted as information drains, and they may occasionally play an important role as prophylactic drainage devices for bleeding, bile leakage, intra-abdominal infections, and massive ascites (13). However, accumulating evidence suggests that prophylactic intra-abdominal drains are unnecessary for both gastrointestinal and liver surgeries because these surgical procedures can be performed safely without drainage (14–17). Drains limit the patient’s activity of daily living scale due to pain at the insertion site and the necessity of a drainage bag; moreover, a longterm placement of the drains can lead to a high risk of intra-abdominal infections by providing a route for ascending infection. In addition, drains after hepatic resection are associated with a significantly longer hospitalization and higher post-operative morbidity in patients with chronic liver diseases (14). Similar results demonstrating a lack of benefit from prophylactic intra-abdominal drains have been reported in patients undergoing cadaveric LT using wholeliver grafts (18–21). However, no studies have investigated on the necessity of prophylactic drains after living donor liver transplantation (LDLT) using split-liver grafts and reduced-liver grafts. LT using split-liver grafts and reducedliver grafts yields a possibility of bleeding and bile leakage from the transection site of graft liver in comparison with whole-liver grafts, and therefore may present a high risk of post-transplant intra-abdominal infections. Therefore, prophylactic intra-abdominal drains and ascitic examinations may be important tools for early diagnosis and treatment of bleeding, bile leakage, and intra-abdominal infections after LDLT. Only our group has previously reported ascitic data after LDLT (13), but there have been no widely agreed-upon methods for the efficacy and management of prophylactic post-transplant intra-abdominal drains. Therefore, the present retrospective study investigated whether ascitic data at POD 5 after LDLT can predict intraabdominal infections and on the post-transplant management of prophylactic drains. 596
Patients and methods Patients Between March 2008 and March 2013, 90 LDLTs were performed on pediatric patients using grafts associated with left lobe at the Department of Transplant Surgery, Jichi Medical University, Japan. The baseline demographic data pertaining to the recipients and grafts are presented in Table 1. Seventeen patients were excluded because insufficient data were collected at POD5 in 12 patients and urgent laparotomy or endovascular treatment until POD5 was undergone due to intra-abdominal bleeding in three patients (POD1, POD2, and POD2), hepatic arterial complication in one (POD4), and abdominal closure in one (POD5). Therefore, all of the recipients exhibited no specific complications (including surgical treatment) until POD5 in this study.
Immunosuppressive therapy Tac and MP were used as the standard post-operative immunosuppressive therapy. The target trough level of Tac was 15–20 ng/mL during the first week, 8–12 ng/mL during the first month, 5–8 ng/mL during the first six months, 3– 5 ng/mL during the first year, and 2–4 ng/mL thereafter. MP was administered at an initial dose of 20 mg/kg intravenously on the morning of the operation and before graft reperfusion. The MP dose was thereafter decreased gradually to three mg/kg/day on POD1, 0.5 mg/kg/day on POD7, and 0.25 mg/kg/day at one month post-LDLT, and was discontinued within one year, except in patients for whom immunosuppression could not be maintained at the lower dose. Mycophenolate mofetil was used when more potent immunosuppression was required, for example, in ABO-incompatible recipients older than five yr, in patients with steroid-resistant acute rejection episodes, and in patients with liver dysfunction after the cessation of MP therapy. Table 1. Demographic data of recipients and grafts Recipient and graft characteristics Gender Age (months) Body weight (kg) Original disease
PELD or MELD ABO compatibility Graft type GV/SLV (%) Operation time Cold ischemic time Warm ischemic time Blood loss volume (mL/kg) Transfusion volume (mL/kg)
Male 41, female 49 17 (0–234) 9.5 (2.6–53.7) Biliary atresia 66, ornithine transcarbamylase deficiency 10, fulminant hepatic failure 3, congenital extrahepatic portosystemic shunt 3, Alagille syndrome 2, graft hepatic failure 2, primary sclerosing cholangitis 2, citrullinemia 1, methylmalonic acidemia 1 9.4 ( 10.9–32.9) Identical 54, compatible 21, incompatible 15 Left lateral segment 60, left lobe 21, S2 monosegment 2, left lobe+caudate lobe 4 71.0 (33.0–130.0) 12 h 30 min (seven h two min–25 h 56 min) one h 54 min (46 min–six h 45 min) 40 min (24 min–two h two min) 75.7 (1.6–589.1) 98.7 (6.3–635.3)
PELD, pediatric end-stage liver disease; MELD, model for end-stage liver disease; GV/SLV, ratio of graft volume to standard liver volume.
Sanada et al.
various recipient and graft variables on intraabdominal infections was assessed, and the results are summarized in Table 3. The univariate analysis revealed that left lobe and S2 monosegment grafts, graft volume to standard liver volume ratio, ≥12 h 30 min of operation time, and ≥two h of cold ischemic time (p < 0.001, p = 0.043, p = 0.026, and p = 0.058, respectively) were independent risk factors for intraabdominal infections. The multivariate analysis revealed that left lobe and S2 monosegment grafts were an independent risk factor for intraabdominal infections (p = 0.006, odds ratio 10.10, and 95% confidence interval [CI] 1.91– 53.80) (Table 4). Ascitic examinations at POD5 in the patients with posttransplant intra-abdominal infections
The patients with post-transplant intra-abdominal infections had significantly higher acsitic neutrophil count, total bilirubin, and LDH levels at POD5 in comparison with patients without infections (p = 0.018, p = 0.018, and p = 0.001, respectively) (Fig. 1a–c). Consequent to ROC analysis of ascitic neutrophil count, total bilirubin, and LDH levels at POD5 in post-transplant intra-abdominal infections, the recommended cutoff value for diagnosing intra-abdominal infections was set at 624 per mm3, 2.3 mg/dL, and 536 IU/L, respectively. The ROC analysis of the ascitic neutrophil count had a sensitivity of 63.6%, specificity of 77.9%, AUC of 0.668, and 95% CI of 0.528–0.813 (Fig. 2a). The ROC analysis of the ascitic total bilirubin level had a sensitivity of 81.0%, specificity of 51.5%, the AUC of 0.645, and 95% CI of 0.500–0.791 (Fig. 2b). The ROC analysis of the ascitic LDH level had a sensitivity of 61.9%, specificity of 87.9%, the AUC of 0.748, and 95% CI of 0.618–0.877 (Fig. 2c).
Table 4. Risk factors of intra-abdominal infections after living donor liver transplantation: multivariate analysis Variables
Odds ratio
95% CI
p-value
Graft type Others vs. left lateral segment graft GV/SLV
10.10
1.91–53.80
0.006
20.50
0.36–1170.00
0.143
2.38
0.54–10.50
0.251
3.09
0.82–11.7
0.097
Operation time ≥12 h 30 min vs.
