Application of Various Surgical Techniques in Orthotopic Liver Transplantation: a 10-year Experience at a Single Center

Background Liver transplantation (LT) is considered the only curative treatment for end-stage liver disease (ESLD), and the surgical techniques of LT have continually evolved and have been modied. In this study, we prospectively analyzed a single-center case series in our center and compared the advantages and disadvantages of each method.


Background
LT is considered the only curative treatment for ESLD 1 . Currently, LT survival within one year can reach approximately 90% with progressive improvement in surgical methods, immunosuppressants and sepsis controls 2 .
Furthermore, the surgical techniques of LT have continually evolved and have been modi ed 3  were completely blocked, and the retrohepatic inferior vena cava (RHIVC) was removed as a part of the diseased liver; it was named classic OLT 4 . However, the hemodynamics in the anhepatic phase is not stable, and re ux of the renal vein is blocked, causing renal dysfunction 5 . In 1989, Tzakis et al. 6 described a novel technique called classic piggy-back (PB) LT. In this technique, the PHVC is preserved, the SHVC of the donor liver is anastomosed with a common opening formed by the left and middle hepatic veins of the recipient liver, and the IHVC of the donor liver is ligated. The ow during the anhepatic period and hemodynamic stability are maintained during the operation.
The technique can also reduce the occurrence of renal failure after liver transplantation and is more bene cial to patients with cardiac insu ciency or a poor general condition. However, speci c complications related to PB, such as out ow obstruction, can cause liver congestion, swelling and even delayed graft function (DGF) or transplanted liver failure. In 1992, Belghiti et al. 7 developed the modi ed piggy-back (MPB) technique in which side-to-side cavocaval anastomosis is performed at the anterior face of the recipient RHIVC to minimize out ow obstruction.
Wu. et al. 8 also reported another MPB technique, which has the advantages of simplifying the steps of hepatectomy and separation, and the anastomosis of the vena cava is large, thereby avoiding out ow tract obstruction. However, this operation requires complete blockade of the inferior vena cava, and the anhepatic period is long, thereby leading to intraoperative hemodynamic instability and renal dysfunction.
The PB and MPB techniques are limited when the caudate lobe of the recipient liver is hypertrophied 9 . Particularly, in some patients with Budd-Chiari syndrome, the RHIVC is surrounded in the caudate lobe of the liver, making it di cult to preserve the inferior vena cava during hepatectomy 10 . Additionally, in patients with large liver cancers, the RHIVC tends to be resected, and classic LT is recommended 11 . To simplify the operation process and shorten the anhepatic period, we developed a modi ed classic LT technique 12 . In this technique, it is not necessary to free the IVC and dissect the posterior space of the IVC and right adrenal vein. The IVC is clamped directly from front to back and repaired after resection of the diseased liver. Next, it is sutured continually from front to back. Compared with conventional classic LT, it is easy to expose and convenient for the surgeon to anastomose.
In general, the selection of surgical techniques is dependent on the patient's conditions. In this study, we prospectively analyzed a single-center case series (1,029 cases) at our center and compared the advantages and disadvantages of each method.

Methods
In total, 1,029 patients who had undergone OLT at our department from January 2009 to December 2019 were enrolled in the study. The recipient pretransplantation data, including demographics, the results of preoperative laboratory examinations, model for end-stage live disease (MELD) score, Child-Pugh score, history, and diagnosis were collected. The perioperative data, including the anhepatic time, cold ischemia time, duration of operation, blood loss and transfusion of red blood cells (RBCs), fresh frozen plasma (FFP), and platelets, were recorded.
Additionally, postoperative surgical complications, length of stay in the intensive care unit (ICU), and the mortality rates and cause of death within the postoperative 30 days were assessed and analyzed.

Modi ed classic technique
This technique is an improvement of classical orthotopic LT. First, the common bile duct, hepatic artery and portal vein were dissociated in turn for dissection of the rst hepatic hilum. Thereafter, the left triangular ligament, hepatogastric ligament and right triangular ligament were separated and the RHIVC was dissociated. The differences from the conventional method are that dissociating the IVC is unnecessary and the blocking direction was from front to back. The donor's and recipient's IVC were xed together at 0 and 6 o'clock using two 4-0 Prolene lines. The suture was continuous from back to front. The patients who had undergone this technique were enrolled in group A.

MPB technique
The dissection of the rst hepatic hilum and peri-hepatic ligaments was the same as that in the convention technique. Subsequently, the SIVC and PIVC were dissociated and then immediately blocked. The diseased liver was resected close to the anterior wall of the IVC, and all short hepatic veins were ligated. Thereafter, the openings of the three hepatic veins were cut and reshaped, and the anterior wall of the IVC was cut longitudinally to form a large triangular outlet. Anastomosis was performed between the outlet of the recipient IVC and matched the opening of the SVC in the donor liver. The IHVC of the donor liver was ligated before reperfusion. The patients who had undergone this technique were enrolled in group B.

PB technique
This technique was described as the standard back-table procedure for the liver. The dissection of the rst hepatic hilum and peri-hepatic ligaments was the same as that in the conventional technique. The short hepatic veins in the third hepatic hilum were ligated so the right, left and middle hepatic veins (RHV, LHV and MHV) were exposed. The diseased liver was then resected after ligation of the RHV and blockade of the LHV as well as the MHV.
Anastomosis was performed between the recipient SVC and the reshaped common trunk of the LHV and MHV of the donor liver. During the entire operation, the blood ow in the IVC was maintained. The patients who had undergone this technique were enrolled in group C.

Postoperative management
Basiliximab was used for induction during the operation and postoperative day (POD) 4. The immunosuppressive regimen was tacrolimus (Tac) and mycophenolate mofetil (MMF). Corticosteroid therapy was not included in the routine regimen because of its side effects. The blood concentration of Tac was controlled at 8 to 15 ng/ml in the early stage after the operation. Routine Doppler ultrasound of the liver graft blood ow and biliary tract was performed once every 2 days for 7 days. Thereafter, imaging studies were performed based on the patients' clinical status or laboratory ndings.

Statistical analysis
All the statistical analyses of the data were performed using SPSS version 26.0. All the data were expressed as means ± standard deviation or numbers and percentages of patients. For comparison between groups, chi-square and Fisher's exact tests were performed for frequencies and continuous data, respectively. The Cox proportional hazards model was used for multivariate analysis. A P-value<0.05 was considered to be signi cant.

Results
We performed 1,029 OLTs at our center from January 2009 to December 2019. Three hundred eighteen patients had undergone the modi ed classic technique (group A), 592 had undergone the modi ed MPB technique (group B), and 119 had undergone the PB technique (group C). The baseline data (gender, age, height and weight) of the patients among the groups were similar ( Table 1). The male to female ratio was 6.74, and the mean age was 48.48 ± 0.40 years. Patients in group C were in poorer general condition with higher creatinine, total bilirubin and PT-INR (P = 0.029, 0.011 and 0.026, respectively), and other indexes were similar. The mean MELD score and Child-Pugh score were 17.02 ± 0.30 and 8.44 ± 0.07, respectively. These values were signi cantly higher in group C. Two hundred ninety-ve patients had a previous abdominal surgery, and the proportion was higher in group B (P = 0.017). Seven hundred forty-four patients were diagnosed with cirrhosis, and 594 were diagnosed with tumors. Cirrhosis was the most prevalent condition in group B, while tumors tended to be more common in group A (P = 0.023 and<0.001, respectively). We also noted a signi cant difference in intraoperative transfusion such that patients in group B had a larger transfusion volume of platelets (P = 0.029). The estimated blood loss and transfusion of FFP and RBC were similar in all the groups ( Table 2). followed by renal failure (n = 46, 4.5%) and early allograft dysfunction (n = 35, 3.4%). Notably, 3 patients had out ow obstruction (group B: n = 2, 0.33%; group C: n = 1, 0.84%; P = 0.010). No signi cantly differences were observed in other complications (Table 3). In all the patients, the mortality rate was 7.9% within 30 days and 11.7% within 90 days ( Table 3). The mortality rates within 30 days in groups A, B and C were 6.9%, 8.6% and 7.6%, respectively. The mortality rates within 90 days in groups A, B and C were 10.0%, 12.7% and 11.8%, respectively. The most frequent causes of death were multiple organ dysfunction syndrome (MODS) (within 30 days: n = 56, 68.3%; within 90 days: n = 84, 69.4%) in all the patients (Table 4). Vena cava obstruction was signi cantly different in group C (P = 0.016 and 0.021, respectively).
In multivariable analysis, we found that abdominal bleeding, portal vein thrombosis (PVT), early allograft dysfunction (EAD) and acute kidney injury (AKI) were independent factors impacting survival within 30 days, and respiratory diseases, abdominal bleeding, hepatic artery thrombosis (HAT), PVT, primary nonfunction (PNF), EAD, AKI were independent factors impacting survival within 90 days (Table 5).   17 . Our center developed a modi ed classic LT to simplify the operation process, shorten the anhepatic time and ensure radical treatment. Zoltan et al. 18 reported that 50% of PB techniques and 40.5% classic techniques were performed as standard techniques in European transplantation centers. Therefore, a reasonable operation technique based on the patient's condition is of great signi cance to improve the prognosis of the patients. In this study, we prospectively analyzed the data of patients who had undergone LT at our center and compared the advantages and disadvantages of each method.
Many patients are diagnosed with liver cirrhosis in China, and the annual incidence is 2%~10% 19 . The annual incidence of decompensated cirrhosis progressed to decompensated cirrhosis is 3-5%, and the 5-year survival rate of decompensated cirrhosis is 14-35%. The annual incidence of HCC in patients with liver cirrhosis is 3-6% 20 . In our study, tumors and cirrhosis tended to be the most prevalent conditions in groups A and C, respectively. MPB would be the optimal choice for patients with nontumor-related ESLD. A large proportion of the tumors in the advanced-stage patients tend to undergo surgical resection or conservative treatment due to economic or ideological reasons, even if the tumors are detected early. LT would be considered only when other treatments were ineffective or if the tumor progressed. Thus, the modi ed classic OLT would be more radical.
The mean operation duration at our center was 511.94 ± 38.22 minutes, and no signi cant difference was found among the groups. Compared with previous national reports (range: 320-708 minutes) 21-23 , our data for the MPB technique (548.48 ± 66.30 minutes) in group B were consistent with the international data. The mean duration in our cases was 461.33 ± 6.40 and 465.43 ± 10.74 minutes in groups A and C, respectively. This result is acceptable compared with that in the randomized trial conducted by Cavallari et al. 24 in which the operative time was 506 ± 85 minutes and 462 ± 87 minutes in the PB and conventional groups, respectively. A larger proportion of patients with previous abdominal surgery in group B may explain the longer operative time. The cold ischemia time is an inherent and unavoidable factor in LT. Egea-Guerrero et al. 25 reported that a long ischemia time ( 6 hours) leads to a higher rate of complications after LT. The mean cold ischemia time in this study was 436.65 minutes and comparable with that in previous reports 26 . The longer cold ischemia time in group B may also be due to the history of abdominal surgery. The anhepatic phase is de ned as the time from the dissection of the recipient liver to reperfusion of the graft. The length of the anhepatic time is variable, depending on the surgical technique. Stuart et al. 27 concluded that the anhepatic phase would increase blood loss because of the absence of hepatic synthesis and clearance. Alexander et al. 28 revealed that a long anhepatic phase duration (> 100 minutes) is an independent risk factor for graft dysfunction in LT. In his study, the median anhepatic time was 71 minutes. In our study, the median and mean anhepatic times were 51 and 54.99 ± 0.66 minutes, respectively, and it was similar in all the groups (P = 0.775). This result may suggest that the modi ed classic LT technique could achieve comparable anhepatic times as PB or MPB because of its simpli ed procedure compared with the classic technique.
We previously reported a new technique to ensure the complete avoidance of ischemia injury during transplantation, de ned as ischemia-free liver transplantation (IFLT) 29 . The donor liver is procured, preserved, and implanted under continuous NMP without a cold perfusion process. The cold ischemia time is theoretically shortened even to 0. The results showed that it had obvious advantages in the recovery of allograft function and reduction of complications compared with the conventional procedure. A randomized clinical trial (RCT) was also performed to con rm its feasibility (ChiCTR1900021158).
According to our study, the median intraoperative blood loss, transfusion of RBCs, and FFP were 1500, 1000 and 1600 ml, respectively, and they were similar in all the groups. A larger transfusion volume of platelets was performed in group B and was within an acceptable range compared with the experience at other centers 30,31 . In a cohort study of over 5,000 patients in France, Savier et al. 32 found that the median ICU duration after LT was 8 days (5 ~ 15 days). Arianeb et al. 23 reported a 14-day ICU and IMC stay in their experience of 500 LTs using the MPB technique. In our study, the median and mean ICU stay time were 35 and 78.62 ± 6.20 hours, respectively, markedly shorter than previous experiences. Our center promoted the concept of 'enhanced recovery after surgery' (ERAS) to achieve early extubation, early functional exercise, and a shortened length of stay at the hospital 33 .
Complications in the early postoperative period are important concerns in LT. In our study, 6.0% of the patients had abdominal bleeding, and it was the most frequent postoperative complication. Abdominal bleeding is related to the preoperative patients' condition, operation techniques and postoperative anticoagulation 34,35 . Nobuhisa et al. 36 reported that a higher MELD score, intraoperative blood loss and low brinogen value increase the incidence of postoperative abdominal bleeding. Our result revealed that it was not signi cantly different among the different techniques.
AKI was another prevalent complication, and its incidence was 4.5% in our study. The incidence of AKI in previous reports ranged from 0-16.7% 37,38 . AKI was mainly caused by hemodynamic instability, IVC blockade, and severe intraoperative blood loss. Hesse et al. 37 showed that the incidence of postoperative renal dysfunction was signi cantly lower in the MPB group. However, no signi cant difference was found in the incidence of renal failure among the different groups in our study (4.7%, 4.4% and 4.2% in groups A, B and C, respectively; P = 0.964). The cause may be that, although the IVC was blocked during the operation in group A, the operation process was simpli ed and the operative time was shortened, leading to little effect on the hemodynamic stability 39 .
Notably, out ow obstruction occurred in both groups B and C, with incidences of 0.16% and 1.68%, respectively.
Joel et al. 40 43 . Additionally, they also mentioned that patients with PVT had a higher mortality rate of 12.4 17.7% 44,45 . HAT and PNF were also the main causes of death for the reported patients according to the ndings of Khan et al. 46 . In multivariable analysis for our cases, we also found the above complications were the independent factors impacting survival.

Conclusion
Different surgical techniques have speci c advantages and disadvantages. MPB or PB would be the optimal choice for patients with nontumor-related ESLD. For patients with advanced tumors, the modi ed classic OLT would be more radical. Our modi ed classic technique could achieve the same operation duration, including the cold ischemia time and total operative time, as MPB or PB and was proven to be as feasible as the classic method. A lower incidence of AKI in every group proved that hemodynamic stability should be maintained during the operation. Out ow obstruction persisted in MPB or PB; thus, the technique should be modi ed such that the incision on the IVC should have su cient distance from the hepatic veins to avoid obstruction of hepatic venous out ow. At our center, most of the deaths were not technique-related; thus, preoperative general condition adjustment and postoperative management need to be further innovated to reduce mortality. IFLT may be the mainstream technique in the future because of its advantages. Therefore, a reasonable operation technique based on the patient's condition to ensure the stability of hemodynamics during the operation is of considerable signi cance to improve the prognosis. Animal Trials of the First A liated Hospital of Sun Yat-sen University on human experimentation (institutional and national) and with the Helsinki Declaration of 1964 and later versions. All the organs used in our study were from organ donation, and none was from executed prisoners. Written informed consent was obtained from all the patients before they were included in the study.

Consent for publication
Not applicable.