- Email: [email protected]
Intraoperative assessment of hepatic venous congestion with direct clamping of the hepatic vein trunk for living donor liver transplantation
Intraoperative Assessment of Hepatic Venous Congestion With Direct Clamping of the Hepatic Vein Trunk for Living Donor Liver Transplantation S. Hwang, S.G. Lee, K.H. Kim, K.M. Park, Y.J. Lee, C.S. Ahn, D.B. Moon, T.Y. Ha, S.H. Cho, and K.B. Oh ABSTRACT We devised a hepatic vein clamping method to assess the amount of hepatic venous congestion (HVC) before liver transection. From February 2003 to May 2003, this method was applied to 5 of 58 living donor livers especially to assess donor safety. The left portal vein and proper hepatic artery as well as the middle hepatic vein (MHV)–left hepatic vein (LHV) trunk were clamped simultaneously to assess the HVC in the remnant right lobe before performing extended left lobectomy. As three donors demonstrated the extent of the HVC equivalent to about 40% of the right lobe volume (RLV), their operations proceeded according to the preoperative plan. The territory of HVC after liver transection was the same as that observed with direct clamping of the hepatic vein. However, one donor showed massive HVC more than 50% of RLV and the operative plan was adjusted to harvest only the left lobe without the MHV trunk for donor safety. To assess the HVC in the remnant left lobe, the isolated LHV trunk was occluded after clamping the donor’s proper hepatic artery. The whole left lobe except for a small area at the anterior portion of the medial segment became discolored on LHV clamping: the opposite demarcation appeared on MHV clamping. The amount of HVC was so small that we harvested the right lobe with the MHV trunk. All donors and recipients recovered uneventfully. We believe that this direct clamping method makes the assessment of HVC feasible before parenchymal transection of a donor liver.
D
URING PARENCHYMAL transection of living donor livers, some large tributaries from the middle hepatic vein (MHV) are severed, because of differences between the inflow and the outflow vascular anatomies. Since deprivation of hepatic venous drainage leads to congestion and dysfunction of the veno-occlusive parenchyma, quantitative assessment of hepatic venous congestion (HVC) has been performed during living donor hepatectomy, especially in adult-to-adult living donor liver transplantation (LDLT) using a right lobe graft. However, before hepatic transection, it has been difficult to exactly evaluate the amount of HVC. To quantitatively predict the HVC in the donor right lobe, we presented a preoperative measurement method of hepatic vein anatomy using donor computed tomogram (CT).1 This method was primarily intended to determine whether to reconstruct the middle hepatic vein tributaries of the right lobe grafts.2 It has become an essential part of living donor assessment in our
institution since the completion of the study in late 2002. On the other hand, there are some other situations that affect whether to harvest the middle hepatic vein (MHV) trunk in view of donor safety, because massive HVC must be detrimental to recovery of both donor remnant livers and hepatic grafts. Such a decision is directly related to living donor safety; it is prudent that accurate data be obtained on the amount of HVC as well as the size of the remnant liver. This information should also be available before parenchymal transection. However, no truly reliable method is available to assess HVC, CT-based predictions of HVC From the Division of Hepatopancreatobiliary Surgery and Liver Transplantation, Department of Surgery, Asan Medical Center, Ulsan University College of Medicine, Seoul, Korea. Address reprint requests to: SungGyu Lee, MD, Department of Surgery, Asan Medical Center, 388-1 Poongnap-Dong, SongpaGu, Seoul 138-736, Korea. E-mail: [email protected]
0041-1345/04/$–see front matter doi:10.1016/j.transproceed.2004.05.018
© 2004 by Elsevier Inc. All rights reserved. 360 Park Avenue South, New York, NY 10010-1710
1462
Transplantation Proceedings, 36, 1462–1465 (2004)
HEPATIC VENOUS CONGESTION
1463
Fig 1. Procedures of direct clamping of the hepatic vein trunk to induce hepatic venous congestion. (A) The left portal vein and proper hepatic artery were separately occluded by two sets of bulldog clamps. The left portal vein was also encircled by a rubber vessel loop. (B) The common trunk of the middle-left hepatic vein was concurrently clamped. (C) Hepatic venous congestion emerged at the anterior segment portion of the right liver. (D) The width of hepatic venous congestion was measured to perform CT volumetry. The amount of hepatic venous congestion was calculated as 39% of the right lobe volume.
showed not so small prediction errors when applied to livers with massive HVC. As direct clamping of a hepatic vein trunk simulates the state of HVC in our previous empirical trials, we adopted this method to assess intraoperatively HVC. This method was applied to assess HVC in the right lobe as well as in the left lobe to determine whether to harvest the MHV trunk before parenchymal transection of a donor liver. MATERIALS AND METHODS From February 2003 to May 2003, we performed 48 cases of adult-to-adult LDLT including 10 cases of dual LDLT.2 We used CT-based preoperative assessment of HVC in the donor right lobe1 to arbitrarily classify donors as massive HVC group (HVC larger than 40% of the right lobe volume (RLV, n ⫽ 14) and nonmassive HVC groups (HVC not larger than 40% of RLV, n ⫽ 44). The donated segments included: right lobe without MHV trunk (n ⫽ 26), right lobe with MHV trunk (n ⫽ 1), left lobe with MHV trunk (n ⫽ 27), left lobe without MHV trunk (extended left lateral segment, n ⫽ 1), left lateral segment (n ⫽ 2), and right posterior segment (n ⫽ 1). Four donors of a left lobe graft and one donor of a right lobe graft in the massive HVC group underwent intraoperative assessment of HVC by directly clamping the hepatic vein trunk. This procedure was empirically indicated when the remnant
right liver mass (except for the HVC portion) was expected to be less than 35% to 40% of total liver volume after procurement of the left lobe with the MHV trunk, or when harvesting a right lobe with a MHV trunk. For assessment of HVC in the remnant right lobe before ongoing left lobectomy, we clamped the left portal vein and proper hepatic artery to induce solitary portal perfusion to the right lobe (Fig 1A). Thereafter, we occluded the common trunk of MHV–left hepatic vein (LHV) with a set of angled vascular clamps (Fig 1B). These procedures simulated the complex state of MHV occlusion following parenchymal transection of the liver and concurrent clamping of the right hepatic artery and total occlusion of the left lobe. The territory of HVC in the right lobe was defined as an intervening portion of intermediate discoloration between the totally darkened left lobe and normal-colored posterior portion of the right lobe (Figs 1C, 2). For assessment of HVC in the left lobe before ongoing procurement of right lobe and MHV trunk, we occluded only the isolated LHV trunk after clamping of the proper or left hepatic artery (Fig 3). This procedure also induced surface discoloration equivalent to the territory of LHV drainage in the left lobe. The counter portion to the discolored area in the medial segment (S4) became the territory of HVC in the remnant left lobe. The direct or counterdiscoloration on the liver surface was measured (Fig 1D) and transferred to the preoperative donor liver CT,3 to obtain volumetry of the corresponding area instantly in the operation
1464
HWANG, LEE, KIM ET AL
Fig 2. Discoloration of the donor liver surface after concurrent clamping of the left portal vein, proper hepatic artery, and middle-left hepatic vein common trunk. The intervening zone of intermediate discoloration between the darkened left lobe and normal-colored posterior portion of the right lobe indicates the territory of hepatic venous congestion induced by deprivation of middle hepatic vein drainage in the right lobe. The amount of discoloration was calculated as 42% of the right lobe volume by using computed tomogram volumetry. room using PACS (Picture Archiving and Communication System, Petavision, Hyundai Information Technology Co, Seoul, Korea).
RESULTS
In four donors of left lobe graft, the territory of HVC in the right lobe was delineated as an intervening discoloration area after 2 to 3 minutes occlusion of the hepatic inflow and outflow. In three donors among them, the amount of HVC was estimated to be 35%, 39%, and 42% of RLV (Figs 1C, 2), which was not so different from the CT-based preoperative prediction. The remnant livers (except for the HVC portion) were calculated to be nearly 40% of the whole liver mass. This verification process for HVC led to combined harvesting of a left lobe graft and MHV trunk without concern for donor safety. The territory of HVC described by direct hepatic vein clamping method was exactly matched to that following parenchymal transection. However, in one donor who showed a CT-predicted HVC of 47% of RLV, the direct clamping method suggested that the discolored area of HVC was estimated at more than 50% of RLV. The remnant liver (except for the HVC portion) would be less than 35% of the whole hepatic mass. After considering the donor age of 45 years, the dark mottling of the liver surface and the apparent blunting of the liver edges, mild steatosis, and presence of two elongated left hepatic vein tributaries to S4, we altered the resection from a left lobe with MHV trunk to a left lobe without an MHV trunk for the safety of donor. The graft was implanted as one transplant with a dual LDLT. Follow-up studies with 7-day and 14-day CT scans demon-
strated nearly no perfusion abnormality in this graft liver without an MHV trunk. This method was also applied to one donor of a right lobe, in whom the MHV trunk was dissected from the LHV trunk. After 2 minutes of clamping of the proper hepatic artery and an isolated MHV trunk, a small triangular area of discoloration emerged at the anterior portion of S4 beside the anterior segment, which was equivalent to 6% of the whole left lobe volume. Another 2 minutes of clamping of the LHV trunk revealed darkening of the whole left lobe except for the anterior portion of S4. The liver resection rate was equivalent to 63% of the total liver volume when performing right lobectomy with MHV procurement and there was a thick umbilical vein with several branches to S4, so we believed that the extent of donor hepatectomy was safe, using the right lobe and MHV trunk. A small portion of attenuation change was detected at the anterior portion of S4 in the 7-day donor CT, but it disappeared upon development of collateral veins on the 2-month CT. All five donors and five recipients recovered uneventfully and survive to date. DISCUSSION
Assessment of HVC has become an essential procedure for LDLT because HVC predisposes to unfavorable recovery of function and impaired regeneration in both donor and recipient.4,5 However, there has been no method to predict the amount of HVC before liver transection. We have developed an HVC prediction method based on the hepatic vein anatomy using a preoperative donor CT: the mean
HEPATIC VENOUS CONGESTION
1465
Fig 3. Discoloration of the donor liver surface after concurrent clamping of the left hepatic artery and left hepatic vein. In this donor, the left hepatic vein was not isolated and the left lateral half of the middle-left hepatic vein common trunk was occluded with a set of vascular clamps. This partial clamping induced faint discoloration in the medial segment, by which the territory of middle hepatic vein drainage could be delineated as a counterdemarcation. The amount of left lobe parenchyma beyond the color demarcation was calculated as 15% of the left lobe volume.
prediction error was 3.8% of RLV.1 However we predicted that the possibility of intrahepatic venous intercommunication would reduce HVC.6 As we had to verify the quantitatively predicted HVC before liver transection, we devised another method to directly assess HVC as a verification process especially for donor safety. The combination of the direct clamping and the CT-based prediction methods allowed us to adjust the extent of donor hepatectomy, in consideration of both donor safety and the functional remnant graft mass. We have generally least 35% of total liver volume in the remnant of 650 living donors to date, resulting in neither donor mortality nor liver functionrelated sequelae. The short duration of hepatic vein clamping in this study was the time for emergence of any noticeable discoloration. Prolonged clamping may make the color change of HVC more apparent. CT volumetry for HVC during the donor operation took only 10 minutes using PACS. We think that this delay in the donor operation is worthwhile. HVC in the right lobe has been intensively investigated, but that in S4 has not been described yet in detail. We observed an impairment of S4 regeneration after procurement of the right lobe and MHV trunk, unless there was an abundance of LHV tributaries to S4.5 We think that this method also can be used to evaluate the hepatic venous anatomy of S4 as in the anterior segment. The perfusion states of HVC in the remnant or grafted livers were demonstrated as attenuation changes in dynamic CT images.7 We think that triphasic CT is the most useful tool to study HVC, because it depicts the intrahepatic
anatomy as well as the dynamic state of parenchymal perfusion. We also think that this method itself without CT scan is applicable to split-liver transplantation because the HVC can be estimated from the size of the HVC-induced discoloration. In conclusion, this initial trial had led us to believe that the direct clamping method makes exact measurements of HVC feasible before parenchymal transection of a donor liver. REFERENCES 1. Hwang S, Lee SG, Park KM, et al: Quantitative prediction of hepatic venous congestion of right lobe grafts using computed tomogram of living donors. Liver Transpl 9:C-70, 2003 (abstract) 2. Lee SG, Park KM, Hwang S, et al: Adult-to-adult living donor liver transplantation at the Asan Medical Center, Korea. Asian J Surg 25:277, 2002 3. Sano K, Makuuchi M, Miki K, et al: Evaluation of hepatic venous congestion: proposed indication criteria for hepatic vein reconstruction. Ann Surg 236:241, 2002 4. Maema A, Imamura H, Takayama T, et al: Impaired volume regeneration of split livers with partial venous, disruption: a latent problem in partial liver transplantation. Transplantation 73:765, 2002 5. Kido M, Ku Y, Fukumoto T, et al: Significant role of middle hepatic vein in remnant liver regeneration of right-lobe living donors. Transplantation 75:1598, 2003 6. Cescon M, Sugawara Y, Sano K, et al: Right liver graft without middle hepatic vein reconstruction from a living donor. Transplantation 73:1164, 2002 7. Murata S, Itai Y, Asato M, et al: Effect of temporary occlusion of the hepatic vein on dual blood supply in the liver: evaluation with spiral CT. Radiology 197:351, 1995
D
URING PARENCHYMAL transection of living donor livers, some large tributaries from the middle hepatic vein (MHV) are severed, because of differences between the inflow and the outflow vascular anatomies. Since deprivation of hepatic venous drainage leads to congestion and dysfunction of the veno-occlusive parenchyma, quantitative assessment of hepatic venous congestion (HVC) has been performed during living donor hepatectomy, especially in adult-to-adult living donor liver transplantation (LDLT) using a right lobe graft. However, before hepatic transection, it has been difficult to exactly evaluate the amount of HVC. To quantitatively predict the HVC in the donor right lobe, we presented a preoperative measurement method of hepatic vein anatomy using donor computed tomogram (CT).1 This method was primarily intended to determine whether to reconstruct the middle hepatic vein tributaries of the right lobe grafts.2 It has become an essential part of living donor assessment in our
institution since the completion of the study in late 2002. On the other hand, there are some other situations that affect whether to harvest the middle hepatic vein (MHV) trunk in view of donor safety, because massive HVC must be detrimental to recovery of both donor remnant livers and hepatic grafts. Such a decision is directly related to living donor safety; it is prudent that accurate data be obtained on the amount of HVC as well as the size of the remnant liver. This information should also be available before parenchymal transection. However, no truly reliable method is available to assess HVC, CT-based predictions of HVC From the Division of Hepatopancreatobiliary Surgery and Liver Transplantation, Department of Surgery, Asan Medical Center, Ulsan University College of Medicine, Seoul, Korea. Address reprint requests to: SungGyu Lee, MD, Department of Surgery, Asan Medical Center, 388-1 Poongnap-Dong, SongpaGu, Seoul 138-736, Korea. E-mail: [email protected]
0041-1345/04/$–see front matter doi:10.1016/j.transproceed.2004.05.018
© 2004 by Elsevier Inc. All rights reserved. 360 Park Avenue South, New York, NY 10010-1710
1462
Transplantation Proceedings, 36, 1462–1465 (2004)
HEPATIC VENOUS CONGESTION
1463
Fig 1. Procedures of direct clamping of the hepatic vein trunk to induce hepatic venous congestion. (A) The left portal vein and proper hepatic artery were separately occluded by two sets of bulldog clamps. The left portal vein was also encircled by a rubber vessel loop. (B) The common trunk of the middle-left hepatic vein was concurrently clamped. (C) Hepatic venous congestion emerged at the anterior segment portion of the right liver. (D) The width of hepatic venous congestion was measured to perform CT volumetry. The amount of hepatic venous congestion was calculated as 39% of the right lobe volume.
showed not so small prediction errors when applied to livers with massive HVC. As direct clamping of a hepatic vein trunk simulates the state of HVC in our previous empirical trials, we adopted this method to assess intraoperatively HVC. This method was applied to assess HVC in the right lobe as well as in the left lobe to determine whether to harvest the MHV trunk before parenchymal transection of a donor liver. MATERIALS AND METHODS From February 2003 to May 2003, we performed 48 cases of adult-to-adult LDLT including 10 cases of dual LDLT.2 We used CT-based preoperative assessment of HVC in the donor right lobe1 to arbitrarily classify donors as massive HVC group (HVC larger than 40% of the right lobe volume (RLV, n ⫽ 14) and nonmassive HVC groups (HVC not larger than 40% of RLV, n ⫽ 44). The donated segments included: right lobe without MHV trunk (n ⫽ 26), right lobe with MHV trunk (n ⫽ 1), left lobe with MHV trunk (n ⫽ 27), left lobe without MHV trunk (extended left lateral segment, n ⫽ 1), left lateral segment (n ⫽ 2), and right posterior segment (n ⫽ 1). Four donors of a left lobe graft and one donor of a right lobe graft in the massive HVC group underwent intraoperative assessment of HVC by directly clamping the hepatic vein trunk. This procedure was empirically indicated when the remnant
right liver mass (except for the HVC portion) was expected to be less than 35% to 40% of total liver volume after procurement of the left lobe with the MHV trunk, or when harvesting a right lobe with a MHV trunk. For assessment of HVC in the remnant right lobe before ongoing left lobectomy, we clamped the left portal vein and proper hepatic artery to induce solitary portal perfusion to the right lobe (Fig 1A). Thereafter, we occluded the common trunk of MHV–left hepatic vein (LHV) with a set of angled vascular clamps (Fig 1B). These procedures simulated the complex state of MHV occlusion following parenchymal transection of the liver and concurrent clamping of the right hepatic artery and total occlusion of the left lobe. The territory of HVC in the right lobe was defined as an intervening portion of intermediate discoloration between the totally darkened left lobe and normal-colored posterior portion of the right lobe (Figs 1C, 2). For assessment of HVC in the left lobe before ongoing procurement of right lobe and MHV trunk, we occluded only the isolated LHV trunk after clamping of the proper or left hepatic artery (Fig 3). This procedure also induced surface discoloration equivalent to the territory of LHV drainage in the left lobe. The counter portion to the discolored area in the medial segment (S4) became the territory of HVC in the remnant left lobe. The direct or counterdiscoloration on the liver surface was measured (Fig 1D) and transferred to the preoperative donor liver CT,3 to obtain volumetry of the corresponding area instantly in the operation
1464
HWANG, LEE, KIM ET AL
Fig 2. Discoloration of the donor liver surface after concurrent clamping of the left portal vein, proper hepatic artery, and middle-left hepatic vein common trunk. The intervening zone of intermediate discoloration between the darkened left lobe and normal-colored posterior portion of the right lobe indicates the territory of hepatic venous congestion induced by deprivation of middle hepatic vein drainage in the right lobe. The amount of discoloration was calculated as 42% of the right lobe volume by using computed tomogram volumetry. room using PACS (Picture Archiving and Communication System, Petavision, Hyundai Information Technology Co, Seoul, Korea).
RESULTS
In four donors of left lobe graft, the territory of HVC in the right lobe was delineated as an intervening discoloration area after 2 to 3 minutes occlusion of the hepatic inflow and outflow. In three donors among them, the amount of HVC was estimated to be 35%, 39%, and 42% of RLV (Figs 1C, 2), which was not so different from the CT-based preoperative prediction. The remnant livers (except for the HVC portion) were calculated to be nearly 40% of the whole liver mass. This verification process for HVC led to combined harvesting of a left lobe graft and MHV trunk without concern for donor safety. The territory of HVC described by direct hepatic vein clamping method was exactly matched to that following parenchymal transection. However, in one donor who showed a CT-predicted HVC of 47% of RLV, the direct clamping method suggested that the discolored area of HVC was estimated at more than 50% of RLV. The remnant liver (except for the HVC portion) would be less than 35% of the whole hepatic mass. After considering the donor age of 45 years, the dark mottling of the liver surface and the apparent blunting of the liver edges, mild steatosis, and presence of two elongated left hepatic vein tributaries to S4, we altered the resection from a left lobe with MHV trunk to a left lobe without an MHV trunk for the safety of donor. The graft was implanted as one transplant with a dual LDLT. Follow-up studies with 7-day and 14-day CT scans demon-
strated nearly no perfusion abnormality in this graft liver without an MHV trunk. This method was also applied to one donor of a right lobe, in whom the MHV trunk was dissected from the LHV trunk. After 2 minutes of clamping of the proper hepatic artery and an isolated MHV trunk, a small triangular area of discoloration emerged at the anterior portion of S4 beside the anterior segment, which was equivalent to 6% of the whole left lobe volume. Another 2 minutes of clamping of the LHV trunk revealed darkening of the whole left lobe except for the anterior portion of S4. The liver resection rate was equivalent to 63% of the total liver volume when performing right lobectomy with MHV procurement and there was a thick umbilical vein with several branches to S4, so we believed that the extent of donor hepatectomy was safe, using the right lobe and MHV trunk. A small portion of attenuation change was detected at the anterior portion of S4 in the 7-day donor CT, but it disappeared upon development of collateral veins on the 2-month CT. All five donors and five recipients recovered uneventfully and survive to date. DISCUSSION
Assessment of HVC has become an essential procedure for LDLT because HVC predisposes to unfavorable recovery of function and impaired regeneration in both donor and recipient.4,5 However, there has been no method to predict the amount of HVC before liver transection. We have developed an HVC prediction method based on the hepatic vein anatomy using a preoperative donor CT: the mean
HEPATIC VENOUS CONGESTION
1465
Fig 3. Discoloration of the donor liver surface after concurrent clamping of the left hepatic artery and left hepatic vein. In this donor, the left hepatic vein was not isolated and the left lateral half of the middle-left hepatic vein common trunk was occluded with a set of vascular clamps. This partial clamping induced faint discoloration in the medial segment, by which the territory of middle hepatic vein drainage could be delineated as a counterdemarcation. The amount of left lobe parenchyma beyond the color demarcation was calculated as 15% of the left lobe volume.
prediction error was 3.8% of RLV.1 However we predicted that the possibility of intrahepatic venous intercommunication would reduce HVC.6 As we had to verify the quantitatively predicted HVC before liver transection, we devised another method to directly assess HVC as a verification process especially for donor safety. The combination of the direct clamping and the CT-based prediction methods allowed us to adjust the extent of donor hepatectomy, in consideration of both donor safety and the functional remnant graft mass. We have generally least 35% of total liver volume in the remnant of 650 living donors to date, resulting in neither donor mortality nor liver functionrelated sequelae. The short duration of hepatic vein clamping in this study was the time for emergence of any noticeable discoloration. Prolonged clamping may make the color change of HVC more apparent. CT volumetry for HVC during the donor operation took only 10 minutes using PACS. We think that this delay in the donor operation is worthwhile. HVC in the right lobe has been intensively investigated, but that in S4 has not been described yet in detail. We observed an impairment of S4 regeneration after procurement of the right lobe and MHV trunk, unless there was an abundance of LHV tributaries to S4.5 We think that this method also can be used to evaluate the hepatic venous anatomy of S4 as in the anterior segment. The perfusion states of HVC in the remnant or grafted livers were demonstrated as attenuation changes in dynamic CT images.7 We think that triphasic CT is the most useful tool to study HVC, because it depicts the intrahepatic
anatomy as well as the dynamic state of parenchymal perfusion. We also think that this method itself without CT scan is applicable to split-liver transplantation because the HVC can be estimated from the size of the HVC-induced discoloration. In conclusion, this initial trial had led us to believe that the direct clamping method makes exact measurements of HVC feasible before parenchymal transection of a donor liver. REFERENCES 1. Hwang S, Lee SG, Park KM, et al: Quantitative prediction of hepatic venous congestion of right lobe grafts using computed tomogram of living donors. Liver Transpl 9:C-70, 2003 (abstract) 2. Lee SG, Park KM, Hwang S, et al: Adult-to-adult living donor liver transplantation at the Asan Medical Center, Korea. Asian J Surg 25:277, 2002 3. Sano K, Makuuchi M, Miki K, et al: Evaluation of hepatic venous congestion: proposed indication criteria for hepatic vein reconstruction. Ann Surg 236:241, 2002 4. Maema A, Imamura H, Takayama T, et al: Impaired volume regeneration of split livers with partial venous, disruption: a latent problem in partial liver transplantation. Transplantation 73:765, 2002 5. Kido M, Ku Y, Fukumoto T, et al: Significant role of middle hepatic vein in remnant liver regeneration of right-lobe living donors. Transplantation 75:1598, 2003 6. Cescon M, Sugawara Y, Sano K, et al: Right liver graft without middle hepatic vein reconstruction from a living donor. Transplantation 73:1164, 2002 7. Murata S, Itai Y, Asato M, et al: Effect of temporary occlusion of the hepatic vein on dual blood supply in the liver: evaluation with spiral CT. Radiology 197:351, 1995