This work was presented at a plenary oral session during the 18th Annual International Congress of the International Liver Transplantation Society (San Francisco, CA, May 16-19, 2012).
Early and Long-Term Results of Routine Microsurgical Biliary Reconstruction in Living Donor Liver Transplantation
Article first published online: 12 FEB 2013
Copyright © 2012 American Association for the Study of Liver Diseases
Volume 19, Issue 2, pages 207–214, February 2013
How to Cite
Lin, T.-S., Chen, C.-L., Concejero, A. M., Yap, A. Q., Lin, Y.-H., Liu, C.-Y., Chiang, Y.-C., Wang, C.-C., Wang, S.-H., Lin, C.-C., Yong, C.-C. and Cheng, Y.-F. (2013), Early and Long-Term Results of Routine Microsurgical Biliary Reconstruction in Living Donor Liver Transplantation. Liver Transpl, 19: 207–214. doi: 10.1002/lt.23582
- Issue published online: 12 FEB 2013
- Article first published online: 12 FEB 2013
- Accepted manuscript online: 29 NOV 2012 01:31PM EST
- Manuscript Accepted: 8 NOV 2012
- Manuscript Received: 6 JUL 2012
We describe our early and long-term experience with routine biliary reconstruction via a microsurgical technique in living donor liver transplantation (LDLT). One hundred seventy-seven grafts (including 3 dual grafts) were primarily transplanted into 174 recipients. The minimum follow-up was 44 months. Biliary reconstructions were based on biliary anatomical variations in graft and recipient ducts. The recipient demographics, graft characteristics, types of biliary reconstruction, biliary complications (BCs), and outcomes were evaluated. There were 130 right lobe grafts and 47 left lobe grafts. There were single ducts in 71.8%, 2 ducts in 26.0%, and 3 ducts in 2.3% of the grafts. The complications were not significantly related to the size and number of ducts, the discrepancy between recipient and donor ducts, the recipient age, the ischemia time, or the type of graft. The overall BC rate was 9.6%. The majority of the complications occurred within the first year, and only 1 patient developed a stricture at 20 months. No new complications were noted after 2 years. When the learning-curve phase of the first 15 cases was excluded, the overall BC rate was 6.79%, and the rate of complications requiring interventions was 2.5%. In conclusion, the routine use of microsurgical biliary reconstruction decreases the number of early and long-term anastomotic BCs in LDLT. Liver Transpl 19:207-214, 2013. © 2012 AASLD.
hepatitis B virus
hepatitis C virus
living donor liver transplantation
microsurgical biliary reconstruction
progressive familial intrahepatic cholestasis
percutaneous transhepatic cholangiodrainage.
Biliary reconstruction has always been regarded as the Achilles’ heel of liver transplantation (LT). This contention is particularly evident in reduced size LT, including split LT and living donor liver transplantation (LDLT). At the onset of whole organ LT, the reported morbidity and mortality rates associated with the procedure were marked at 30% to 50% and 25% to 30%, respectively.[1, 2] With experience and advances in surgical techniques,[3-5] immunosuppressive agents, and postoperative care, the rate of biliary complications (BCs) after whole organ LT is now reported to be as low as 10%. However, the BC rate after LDLT has remained high. The incidence of BCs in LDLT patients has been reported to be 16% to 67%.[3-5, 8-14] These BCs have been conspicuously ascribed to technical difficulties due to the small size and multiple ducts, particularly in right lobe liver grafts. To overcome these complexities, our institution has routinely employed the use of microsurgical biliary reconstruction (MBR) in LDLT since 2006. In our previous report, we elucidated the feasibility and aptitude of this technique for significantly lowering the BC rate in LDLT. Here we highlight the refinements made in our MBR technique to lower the rate of BCs according to a classification of the types of biliary reconstruction variations in LDLT with the microsurgical technique, and we describe the early and long-term results with this technique.
PATIENTS AND METHODS
From March 22, 2006 to May 31, 2012, routine MBR was performed in 580 grafts for 577 consecutive LDLT procedures (including 3 dual graft transplants) at Kaohsiung Chang Gung Memorial Hospital, (Kaohsiung, Taiwan). To assess both early and long-term results, we included only 177 grafts (including 3 dual graft transplants) primarily transplanted into 174 LDLT patients before December 31, 2008 for the analysis. The minimum follow-up was 44 months, and 73 recipients had a minimum follow-up of 5 years. The study protocol was approved by the Chang Gung Memorial Hospital Institutional Review Board.
All biliary reconstructions were performed with a microsurgical technique by a single microsurgeon. The classification of biliary reconstructions was based on the number of ducts in the graft, the manner in which these ducts were reconstructed (with or without ductoplasty), and the conduit (recipient duct or jejunum) used to reconstruct the biliary tree.
Ultrasound was initially used to ascertain significant fluid collections and biliary dilatations. Patients with ultrasound findings suggestive of vascular complications or BCs were further evaluated with computed tomography angiography or magnetic resonance imaging. After discharge, Doppler ultrasound examinations were performed regularly for recipients with vascular complications, every 3 months for patients with hepatocellular carcinoma (HCC), and every 6 months for non-HCC cases. Magnetic resonance cholangiography was performed for cases suspicious for BCs.
Early BCs were defined as BCs occurring within the first 12 months after LT. Late BCs were defined as complications occurring after 12 months. Bile leakage was defined as the presence of bile material in a closed suction drain that persisted more than 7 days after transplantation or as the presence of a biloma around the area of the anastomosis. An anastomotic biliary stricture was defined as an intrahepatic biliary dilatation > 3 mm in the presence of a notable anastomotic narrowing, on the basis of symptoms, or on the basis of abnormal liver function tests. Jaundice, fever, or abdominal pain could be present with a bile leak or stricture. A biloma was diagnosed with ultrasound, computed tomography, or magnetic resonance imaging. A stricture was diagnosed with ultrasound or magnetic resonance imaging and was confirmed with percutaneous transhepatic biliary imaging or endoscopic retrograde cholangiography.
BCs, including leaks and strictures, were commonly managed with nonoperative measures, including radiological interventions such as endoscopic treatments and percutaneous transhepatic cholangiodrainage (PTCD).
Modifications in the MBR Technique
The MBR technique is described in a previous report.[15, 16] Briefly, all biliary reconstructions were performed under an operating microscope (Carl Zeiss, Jena, Germany) with a magnification of × 5 to 15. Each anastomosis was performed with 6-0 Prolene sutures (Johnson & Johnson, Somerville, NJ) and a 6-0–gauge cardiovascular point needle. The interrupted-suture technique was used for the posterior wall anastomosis first, and then the continuous-suture and interrupted-tie technique was used for the anterior wall. All suture knots were tied extraluminally.
When the MBR technique was initially applied, the recipient duct was reduced in size (recipient reduction ductoplasty) and fashioned in a way to approximate the opening of the graft duct (if this was required). This reduction ductoplasty was performed by an LT surgeon without the aid of an operating microscope or a surgical loupe. Thereafter, the graft and recipient bile ducts were anastomosed by the microsurgeon under an operating microscope with × 5 to 15 magnification. When the graft had multiple duct openings, the choice of performing graft ductoplasty before the biliary anastomosis or performing 2 or 3 separate anastomoses for biliary reconstruction was left exclusively to the judgment of the microsurgeon. However, this practice resulted in an unacceptably high BC rate (40%) in the first 15 cases. After reviewing the technique, we found that despite the recipient reduction ductoplasty, there remained a significant degree of discrepancy between the recipient and graft ducts. Furthermore, the microsurgeon found it difficult to fix such incongruence after the completion of the anastomosis. We believe that the majority of the initial anastomotic BCs were largely attributable to this factor. Hence, we instituted the following modifications in the technique:
- Recipient reduction ductoplasty, if this was required because of a discrepancy in the graft and recipient ducts, was to be performed by the microsurgeon under an operating microscope.
- Recipient reduction ductoplasty was to be performed after the completion of the anastomosis on the posterior wall and after the laying down of all sutures on the anterior wall of the duct. Moreover, the sutures on the anterior wall of the duct were to be laid down continuously and coursed through the anterior and posterior walls of the remaining length of the recipient duct. This approach precisely approximated the quantity of sutures to securely close the remaining length of the recipient duct (Fig. 1A,B).
- When there were multiple duct openings in the graft, donor ductoplasty was to be performed only if the interductal distance (IDD) between the openings was equal to or less than the diameter of the smaller or smallest duct opening. Separate anastomoses were to be performed if the IDD was more than the diameter of the smaller or smallest duct opening.
- For these technical modifications and refinements, we created a biliary reconstruction classification system to serve as a guideline for choosing what type of reconstruction would best suit a particular situation.
Biliary Reconstruction Classification
Biliary reconstructions were classified according to the number of graft duct openings, the manner in which these ducts were reconstructed (with or without ductoplasty), and the type of conduit used for reconstructing the biliary tree. The conduits used for reconstruction could be either the recipient duct (right or left hepatic duct or common hepatic duct) or the jejunal Roux limb. The sizes of the duct openings and IDD when there were 2 or more graft duct openings were measured with a caliper. The size of the opening created in the jejunal Roux limb was always patterned to that of the hepatic ducts in the graft.
Single Duct Orifice
- In 1-to-1 reconstruction, a single duct opening in the graft was anastomosed to an opening in the recipient duct or jejunum. Reduction ductoplasty was performed when it was necessary.
Two Duct Orifices
- In 2-in-1 reconstruction, 2 duct openings in the graft were connected together by ductoplasty and were reconstructed to an opening in the recipient duct or jejunum. This type of reconstruction was performed when the IDD between the 2 ducts in the graft was equal to or less than the diameter of the smaller duct opening.
- In 2-to-2 unmixed reconstruction, 2 duct openings in the graft were reconstructed separately to 2 openings in the recipient duct or jejunum. This type of reconstruction was performed when the IDD was greater than the diameter of the smaller duct opening.
- In 2-to-2 mixed reconstruction, 2 duct openings in the graft were reconstructed separately with the recipient duct and jejunum. This type of reconstruction was performed when the IDD between the graft ducts was greater than the diameter of the smaller duct opening, there was only 1 bile duct opening in the recipient duct, and its size could accommodate only 1 graft duct opening. The other duct opening was reconstructed with a jejunal Roux limb.
- In 2-to-1 reconstruction, 2 duct openings in the graft were anastomosed separately, but a single large recipient duct was used. The recipient duct was partially sutured in its mid part and was fashioned in such a way to accommodate the 2 graft ducts separately. This type of reconstruction was performed when the IDD between the 2 ducts was greater than the diameter of the smaller duct opening and when there was only 1 opening in the recipient duct with a size big enough to accommodate the 2 graft ducts.
Three Duct Orifices
- In 3-in-1 reconstruction, 3 duct openings in the graft were connected together by ductoplasty and anastomosed to a single opening in the recipient duct or jejunum. This reconstruction was performed when the IDD between the ducts was equal to or less than the diameter of the smallest duct openings.
- In 3-to-3 unmixed reconstruction, 3 duct openings in the graft were reconstructed separately to 3 openings in the recipient duct or jejunum. This was performed when the IDD between adjacent ducts of the 3 ducts was greater than the diameter of the smallest duct opening.
- In 2-in-1 and 1-to-1 unmixed reconstruction, 2 of the 3 graft ducts were ductoplastied and reconstructed to 1 of the openings in the recipient duct or jejunum. The remaining duct was reconstructed to another opening distant from the first anastomosis. This type of reconstruction was performed when 2 of the 3 ducts had an IDD less than or equal to the diameter of the smallest duct opening in the graft and a distant third duct had an IDD greater than the diameter of the smaller duct opening.
- In 2-in-1 and 1-to-1 mixed reconstruction, 2 of the 3 graft ducts were ductoplastied and reconstructed to a single opening in the recipient duct. The remaining duct was reconstructed to a jejunal Roux limb. The conduits (duct and jejunum) could be used interchangeably. This type of reconstruction was performed when there was only 1 duct available to accommodate 1 or 2 duct openings in the graft. It was likewise performed when 2 of the 3 ducts had an IDD less than or equal to the diameter of the smallest duct opening in the graft and a distant third duct had an IDD greater than the diameter of the smaller duct opening.
Means, standard deviations, and ranges were used as measures for continuous variables, and proportions were used to express categorical variables. The Student t test was used for statistical comparisons of means, whereas Fisher's exact test was used for comparisons of proportions. Statistical analyses were performed with commercially available statistical software (Minitab 15, Minitab, Inc., State College, PA). A P value<0.05 was considered significant.
One hundred seventy-seven grafts were transplanted into 174 recipients. Dual graft LT was performed for 3 patients. There were 125 male recipients (71.84%) and 49 female recipients (28.16%). The overall mean age was 42.63±20.82 years (range=4 months to 67 years). Thirty-four recipients (19.54%) were in the pediatric age group (<16 years old) with a mean age of 3.09±3.20 years (range=4 months to 10 years). The mean Model for End-Stage Liver Disease score was 13.28 (range=1–50). The mean Pediatric End-Stage Liver Disease score was 12.47 (range=−7 to 50). The indications for LT were HCC (41.95%), hepatitis B virus (HBV)–related cirrhosis (19.54%), hepatitis C virus (HCV)–related cirrhosis (8.04%), HBV- and HCV-related cirrhosis (1.15%), alcoholic cirrhosis (5.17%), biliary atresia (14.94%), and other (9.19%).
Overall, there were 130 right and 47 left lobe grafts. Single duct opening was noted in 127 (71.75%), two duct openings in 46 (25.99%) and three duct openings in 4 (2.26%) grafts. The mean size of the duct openings was 3.8 mm (range, 1-10 mm). The proportion of multiple duct openings was significantly higher in right lobe grafts (Table 1).
|Total grafts [n (%)]||177 (100)|
|Right lobe grafts||130 (73.4)|
|Left lobe grafts||47 (26.6)|
|Single grafts [n (%)]||171 (96.6)|
|Right lobe grafts||129|
|Left lobe grafts||42|
|Dual grafts [n (%)]||6 (3.4)|
|Right lobe grafts||1|
|Left lobe grafts||5|
|Single duct opening [n (%)]||127 (71.75)|
|Right lobe grafts||83|
|Left lobe grafts||44|
|Two duct openings [n (%)]||46 (25.99)|
|Right lobe grafts||43|
|Left lobe grafts||3|
|Three duct openings [n (%)]||4 (2.26)|
|Right lobe grafts||4|
|Left lobe grafts||0|
|Mean duct size (mm)a||3.8±1.3 (1–10 mm)|
|Cold ischemia time (minutes)a||55.80±77.24 (18–229)|
|Warm ischemia time (minutes)a||50.91±14.39 (23–115)|
|BCs [n/N (%)]||17/177 (9.6)|
|Right lobe grafts||13/130 (10.0)|
|Left lobe grafts||4/47 (8.51)|
MBR Outcomes Based on Classification
Table 2 summarized the biliary reconstructions based on our classification. 1-to-1 reconstruction was performed in 127 (71.75%) [98 (55.37%) with the recipient duct and 29 (16.38%) with the jejunum]. 2-in-1 reconstruction was performed in 26 (14.69%). This group consisted of 23 (12.99%) 2-in-1 duct-to-duct and 3 (1.69%) 2-in-1 duct-to-jejunum reconstructions. 2-to-2 unmixed reconstruction was performed in 19 (10.73%). This group comprised of 18 (10.17%) 2-to-2 duct-to-duct and 1 (0.56%) 2-to-2 duct-to-jejunum reconstructions. 2-to-2 mixed reconstruction was performed in 1 (0.56%). 3-in-1 reconstruction was used in 1 (0.56%) where recipient duct was utilized as conduit. 3-to-3 unmixed reconstruction was performed in 2 (1.13%) where all utilized the recipient duct as conduit. 2-in-1 duct-to-duct and 1-to-1 duct-to-duct unmixed reconstruction was performed in 1 (0.56%).
|Total grafts (n)||177|
|Single duct opening in donor graft [n (%)]||127 (71.75)|
|1-to-1 reconstruction||127 (71.75)|
|1-to-1 duct-to-duct reconstruction||101 (57.06)|
|1-to-1 duct-to-jejunum reconstruction||29 (16.38)|
|Two duct openings in donor graft [n (%)]||46 (25.99)|
|2-in-1 reconstruction||23 (12.99)|
|2-in-1 duct-to-duct reconstruction||20 (11.30)|
|2-in-1 duct-to-jejunum reconstruction||3 (1.69)|
|2-to-1 duct-to-duct reconstruction||0 (0)|
|2-to-2 unmixed reconstruction||19 (10.73)|
|2-to-2 duct-to-duct reconstruction||18 (10.17)|
|2-to-2 duct-to-jejunum reconstruction||1 (0.56)|
|2-to-2 mixed reconstruction||1 (0.60)|
|Three duct openings in donor graft [n (%)]||4 (2.26)|
|3-in-1 duct-to-duct reconstruction||1 (0.56)|
|3-to-3 duct-to-duct unmixed reconstruction||2 (1.13)|
|2-in-1 duct-to-duct and 1-to-1 duct-to- duct unmixed reconstruction||1 (0.56)|
|Total duct-to-duct and duct-to-jejunum anastomoses (n)||202|
|Total duct-to-duct anastomoses [n (%)]||167 (82.67)|
|Total duct-to-jejunum anastomoses [n (%)]||35 (17.33)|
|Total size discrepancy in duct-to-duct anastomoses [n (%)]||118 (70.66)|
|Size of duct discrepancy (mm)a||1.17±0.74 (0.5-4.5)|
Overall, duct-to-duct anastomosis was performed in 167 (82.67%) and duct-to-jejunum Roux limb in 35 (17.33%) biliary reconstructions. Size discrepancy in the graft and recipient ducts was noted in 118 (70.66%) reconstructions. The mean size of these discrepancies was 1.17 mm (range, 0.5-4.5 mm).
The overall number of BCs was 17 (9.6%). There were 16 early BCs (9.0%); these included 11 bile leaks (6.2%) and 5 biliary strictures (2.8%). The complications consisted of bile leaks (n=4), bile leaks with biloma formation (n=7), bile leaks followed by biliary strictures (n=2), and strictures (n=3). The mean time to the detection of bile leakage was 13 days (range=8–17 days). The mean time to the detection of a biloma was 30.9 days (range=10–57 days). The mean time to the detection of strictures was 194.2 days (range=10–600 days). The number of days to the detection of a stricture was skewed by a recipient who developed hepatic artery thrombosis and another recipient who developed a late BC following late-onset portal vein thrombosis. In the 2 recipients who developed bile leakage and then a stricture, the mean time to the detection of leakage was 12 days (range=8–16 days). The strictures followed after a mean interval of 31.5 days (range=16–47 days).
Eight of the 11 patients with bile leaks experienced spontaneous resolution, 2 underwent a revision of the biliary reconstruction, and 1 required pigtail drainage. All 6 patients with biliary strictures underwent percutaneous transhepatic or endoscopic dilatation and stenting. Overall, 9 patients (5.2%) required intervention for BCs. When the learning-curve phase of the first 15 cases was excluded, the BC rate decreased significantly to 6.79 (n=11); furthermore, only 4 patients (2.5%) required interventions. There were no significant differences between right and left lobes. There was no mortality associated with early BCs.
As for late BCs, 1 biliary stricture (0.6%) developed after a late-onset portal vein complication in an adult patient 20 months after LT. No new anastomotic BCs were noted more than 2 years after LT (Table 3). In the long term, there has been no mortality related to BCs since the start of routine MBR in 2006.
|LDLT Case||Diagnosis||Reconstruction||Complications||Interval to Complications||Intervention||Case Range|
|252||Urea cycle defect||Duct-to-duct||Leak||13 days||Revised to Roux-en-Y||First 15 cases|
|256||HBV||Duct-to-duct (2-in-1)||Stricture||25 days||PTCD|
|262||HBV/HCC||Duct-to-duct (2-to-2)||Leak and biloma||30 days||Resolved spontaneously|
|263||HBV/HCC||Duct-to-duct (2-in-1)||Leak/stricture||16/16 days||PTCD|
|264||HBV/HCV/HCC||Duct-to-duct (3-in-1)||Leak/stricture||8/47 days||PTCD|
|265||HCV/HCCa||Duct-to-duct||Leak and biloma||27 days||Pigtail drainage/T-tube|
|269||HCV||Duct-to-duct||Leak and biloma||57 days||Resolved spontaneously||85 cases|
|274||Biliary atresia||Roux-en-Y||Leak and biloma||19 days||Resolved spontaneously|
|280||HBV/HCCa||Duct-to-duct (2-to-2)||Stricture||10 days||PTCD|
|297||PFICb||Duct-to-duct||Stricture||142 days||Revised to Roux-en-Y/PTCD|
|301||HBV||Duct-to-duct||Leak||17 days||Resolved spontaneously|
|328||HCV||Duct-to-duct, Roux-en-Y||Leak and biloma||33 days||Resolved spontaneously|
|341||Alcohol||Duct-to-duct (2-to-2)||Leak||8 days||Resolved spontaneously||174 cases|
|348||HBV/HCC||Duct-to-duct (2-to-2)||Leak and biloma||10 days||Pigtail drainage|
|360||HBV||Duct-to-duct (2-in-1)||Leak||14 days||Resolved spontaneously|
|396||HBV/HCC||Duct-to-duct||Leak and biloma||40 days||Resolved spontaneously|
Comparison of Complications Before and After Modifications in the Technique
There were 167 duct-to-duct and 35 duct-to-jejunum reconstructions. The recipients who underwent duct-to-duct anastomoses were significantly older (50.8±11.2 years vs. 8.9±17.0 years; P <0.001), with more right lobe grafts utilized [127 (88.9%) vs. 3 (8.9%); P<0.001], longer cold (59.4±26.7 min vs. 39.5±22.8 min; P<0.001) and warm ischemia (51.4±13.9 min vs. 45.9±12.3 min; P=0.03) times, and larger duct openings (3.9±1.4 mm vs. 3.1±0.8 mm; P<0.001). Expectedly, duct size (graft-to-recipient) discrepancy occurred only in duct-to-duct reconstructions. The other variables which included frequency of multiple duct openings and Model for End-stage Liver Disease scores were comparable between the two groups. The BC rates were not significantly higher in patients who underwent duct-to-duct reconstruction [15 (10.5%) vs. 2 (5.9%), P=0.53].
There was a notable drop in the number of duct-to-duct BC after the institution of modifications to the technique [from 15 (8.97%) pre-modification to 9 (5.88%) post-modification]. The BC rate between duct-to-duct and duct-to-jejunum anastomoses after the modification remained comparable [9 (7.0%) vs 2 (6.3%), P >0.99].
Difference Between Grafts Containing Single and Multiple Ducts
There were 127 recipients with grafts containing single duct openings (including 3 patients with dual grafts) and 50 recipients with grafts containing multiple duct openings. The mean diameter of the grafts with a single duct opening was significantly larger than the mean diameter of the grafts with multiple duct openings. Excluding the learning curve (first 15 cases), there was no significant difference in the BC rate between recipients with grafts containing a single duct opening and recipients with grafts containing multiple duct openings [7/114 (6.14%) versus 4/45 (8.89%), P=0.50].
Complications related to biliary reconstruction have resulted in a considerable number of graft failures and deaths among LT recipients.[10, 13] Hence, investigators have relentlessly delved into their causes and created several innovations to overcome BCs. Some of these advances include the acquisition of a comprehensive understanding of the liver, the biliary tree, and its blood supply[17-21]; the creation of novel techniques for hepatic dissection and biliary reconstruction[5, 8, 10, 11, 22-24]; the use of T-tubes and stents[8, 25, 26, 28, 29]; and the application of microsurgical techniques to biliary reconstruction.
The scarcity of organs for LT has resulted in the widespread acceptance of partial liver grafts. However, the early use of these grafts resulted in a mounting rate of biliary adversities. The reported rate of BCs from earlier series ranged from 16% to 67%.[3-5, 8-14] The problem has often been attributed to the disparity in the bile duct anatomy of partial liver grafts. In contrast to whole organ LT, the bile duct in reduced size grafts—particularly right lobe grafts—is conspicuously small, and at times, there are multiple ducts.[11, 30] Furthermore, the sizes of the duct openings in reduced grafts are often divergent from (frequently smaller than) the size of the recipient duct. These predicaments add up to difficulties in biliary reconstruction and a greater risk of BC development in LDLT.[3, 8, 11] For this reason, some surgeons have performed duct-to-duct reconstruction in select grafts to secure a single bile duct anastomosis.[12, 31] Such an impasse could result in the exclusion of otherwise suitable living liver donors.
Currently, biliary reconstruction with a microsurgical technique is one of the most important innovations for lowering the number of BCs. This approach has the technical advantage of enhanced visualization of the operative field under magnification, which helps to prevent physical trauma to the bile duct epithelium and allows more precise placement of stitches during the creation of the anastomosis. Our team has used MBR routinely since March 2006. Our report comparing this technique to the conventional method showed that the risk of developing BCs with MBR was reduced after ample experience and refinements of the technique; in contrast, complications with the conventional method increased even after the procedure had been performed for years. The outcomes of this current series demonstrated a further decline in BCs with microsurgical techniques, and the rate at which interventions were required to treat complications was further lowered to 2.5%.
In a recent review of bile duct anastomotic strictures after adult-to-adult LDLT by Chok et al., warm and cold ischemia times, stent use, postoperative acute cellular rejection, and University of Wisconsin solution use were shown by univariate analyses to be factors associated with strictures; according to a multivariate analysis, the cold ischemia time and acute cellular rejection were significant factors. In that series, the number of graft openings and the sizes of the bile ducts were not significant factors for BCs. This may have been because the median size of the smallest bile duct in their series was 5 mm. In comparison, our mean bile duct size was smaller at 3.8 mm. Also, in contrast to other studies,[13, 31] the number and sizes of the duct openings did not prove to be significant factors in BC development in our series. These findings robustly suggest that the microsurgical technique is capable of surmounting the difficulties due to the small size and multiple ducts of reduced size grafts as well as the discrepancies in the duct sizes of the recipient and graft ducts.
However, such improvements can be attained only if MBR is employed along with the technical refinements instituted to prevent BCs. Our findings also show that in conjunction with the use of the microsurgical technique, we have devised a classification scheme and guidelines for comprehensively defining the types of biliary reconstruction to be employed in each particular case. The decision on the way in which a duct should be reconstructed relies on the type of conduit (duct or jejunum), the number of ducts present in the graft, and the IDD when multiple ducts (2 or more) are encountered. The improvement may also be due to switching from a general or LT surgeon to a microsurgeon in addition to the use of a microscope. The differences in the techniques employed by the microsurgeon in reduction ductoplasty and the use of the IDD in determining when to perform ductoplasty could be associated with the decreasing number of BCs. These were crucial when we modified our MBR technique.
Duct-to-duct anastomosis was preferred in the majority of the cases. However, Roux-en-Y jejunal reconstruction was performed in patients with diseased or absent extrahepatic bile ducts (eg, patients with biliary atresia or primary sclerosing cholangitis) and when the recipient duct was unfit for reconstruction (ie, it was devascularized and short). When there were multiple duct openings in a graft, the decision to perform ductoplasty or make a separate anastomosis relied utterly on the IDD. Ductoplasty was performed when the IDD between the ducts was equal to or less than the diameter of the smaller opening. On the other hand, separate duct anastomoses were performed when the IDD was greater than the diameter of the smaller opening. These techniques are to be performed by a microsurgeon.
Our technical experience has shown that the rate of BCs in LDLT can be reduced remarkably not only by preserving the blood supply of the biliary tree but also by planning the appropriate type of biliary reconstruction and properly aligning the anastomosis of the graft and recipient hepatic ducts. The latter objective is achieved through the modifications that have been instituted in our techniques.
In summary, the routine use of MBR capably surmounts the difficulties due to anatomical variations and size discrepancies between graft and recipient hepatic ducts with excellent outcomes. The routine use of MBR can decrease the number of early and long-term anastomotic BCs in LDLT.