J Hepatobiliary Pancreat Surg (2002) 9:5560

Basic knowledge of interest An anatomical study of short hepatic veins, with special reference to delineation of the caudate lobe for hanging maneuver of the liver without the usual mobilization
Toshio J. Sato1, Ichiro Hirai2, Gen Murakami1, Tetsuhiro Kanamura3, Fumitake Hata4, and Koichi Hirata4
1 2

Department of Anatomy, Sapporo Medical University School of Medicine, South 1 West 17, Chuo-ku, Sapporo 060-8556, Japan First Department of Surgery, Yamagata University School of Medicine, Yamagata, Japan 3 First Department of Surgery, Nara Medical University School of Medicine, Nara, Japan 4 First Department of Surgery, Sapporo Medical University School of Medicine, Sapporo, Japan

Abstract Background/Purpose. The present study was designed to anatomically assess a very recently reported hanging maneuver of the liver without mobilization, in which forceps are inserted blindly between the inferior vena cava (IVC) and liver parenchyma. Methods. We dissected 56 formalin-xed livers (1) to determine whether preservation of the caudate vein (the largest vein draining Spiegels lobe) and inferior right hepatic vein (IRHV) was possible and (2) to identify the territories drained by other, non-preserved short hepatic veins. Results. A potential space for insertion of the forceps was found between the openings of the caudate vein and IRHV; however, if preservation of both veins is absolutely necessary, we recommended protecting the IRHV, such as by taping and retracting it. We classied the other short hepatic veins into two categories, i.e., those draining the left portal vein territory and those draining the right territory. The distributions of the openings of the veins in these territories overlapped. Conclusions. Clear delineation of the left caudate lobe according to the drainage veins appeared to be difcult when the liver was divided along a straight line in front of the IVC. Key words Caudate lobe Short hepatic vein Inferior vena cava Hanging maneuver without mobilization Liver surgery

Introduction Belghiti et al.1 have devised a new maneuver for the surgical removal of the right lobe of the liver (containing a large hepatocellular carcinoma), with preservation of the caudate lobe on the left side. In their procedure, without mobilization of the liver, forceps are inserted into a potential space between the parenchyma of the caudate lobe and the retrohepatic inferior vena cava (IVC) from the caudal side, to the cranial side, guided Offprint requests to: G. Murakami Received: March 22, 2001 / Accepted: August 21, 2001

only by a nger placed in a pocket-like space between the terminal portions of the middle and right hepatic veins (MHV and RHV). Insertion of the forceps is followed by taping, hanging (holding) of the liver, and anterior dissection alongside the MHV. Because of the lack of mobilization, it seems to be a very convenient procedure. Although a diffuse distribution of the openings of veins draining the liver into the IVC was suggested previously (Chang et al.;2 Camargo et al.3), Belghiti et al.1 postulated a longitudinal avascular virtual plane, that would allow the blind insertion of forceps among the venous openings in the inside of the IVC. Although, to the best of our knowledge, no anatomical investigations have been carried out since the presentation of their methods, some surgeons who have learned of this maneuver have started to employ it, especially to prepare extended left grafts for living-related liver transplantation. Surgeons are expecting that this method will allow them to include the left half of the caudate lobe with its effective drainage veins in the graft, as reported by Takayama et al.4 with the usual mobilization, but without mobilizing the liver. However, if the maneuver of Belghiti et al.1 is employed, the terminal portion of the caudate vein, i.e., the thick vein(s) draining Spiegels lobe, should lie on the left side of the forceps, and the opening into the IVC of the inferior right hepatic vein (IRHV) (Makuuchi et al.5), if present, should lie on the right side. Is there really an area free of venous openings between these veins? Do the openings of other retrohepatic short hepatic veins lie on the left or the right side of the forceps? To answer these questions, we designed this study to experimentally clarify the topographic relationship between the forceps and various retrohepatic short hepatic veins, including the IRHV and caudate vein. More specically, we used anatomical specimens in an attempt to identify the venous openings that lie to the left and right of the inserted forceps (i.e., the segments or subsegments that the veins drain).

56

T.J. Sato et al.: Anatomy for hanging maneuver without liver mobilization

Materials and methods Fifty-six whole livers, with no evidence of cirrhosis or macroscopic tumors, were obtained from donated cadavers after the cadavers had been used for anatomical dissection by medical or dental students. The liver specimens did not include any that were too thin or that had excess upper protrusion of S8 or any in which the left side of the liver was much smaller than the right side. After we had opened the IVC on its dorsal side, we conrmed that all 56 livers were suitable for investigation of the distribution of venous openings. The precise course of the forceps inserted ventral to the IVC by Belghiti et al.1 was not known, because they inserted the forceps blindly, and because they did not describe the course in sufcient detail. We hypothesized that there were three possible courses, represented by three paths of a curved, metal wire (1 mm in diameter) placed on the dorsal aspect of the liver (Fig. 1), used instead of the forceps ventral to the IVC. The right course extended from the right inferior protrusion of the liver (S6a) to a pocket-like space between the MHV and RHV. The left course followed a curved line between the pocket-like space and the deepest point of the gallbladder bed. The third course, the intermediate course, lay between the left and right courses. After labeling the three hypothetical courses of the forceps on the dorsal aspect of the liver and after depiction of the results, we dissected the specimens to identify the drainage territory of the short retrohepatic hepatic veins. The territories were identied by the portal branch supplying them. Spiegels lobe is usually supplied by portal vein branches of both left portal vein and hilar bifurcation origin.610 However, we decided that if we found a hilar bifurcation branch in this study, we would consider it to be a left branch, because we believe that, in future it will be possible to include the hilar bifurcation branch in the pedicle for some left-lobe liver grafts. We therefore classied the short veins into six categories: (1) caudate veins, one or two veins, over 3 mm in diameter, draining Spiegels lobe; (2) short hepatic veins from the left caudate lobe (SHV-LC), thin drainage veins (13 mm in diameter at their point of entry into the IVC) arising from the left half of the caudate lobe (left portal territory and hilar bifurcation territory); (3) short hepatic veins from the right caudate lobe (SHV-RC), thin veins (13 mm) draining the right half of the caudate lobe; (4) IRHV, one or two thick veins over 5 mm in diameter at their point of entry into the IVC and arising from S6; (5) middle right hepatic vein (MRHV) after Couinaud,11 over 5 mm in diameter, draining S7, including the dorsolateral paracaval area;12 (6) SHV-8, vein(s) over 1 mm in diameter draining S8, including its most posteromedial territory (the so-called PV8c territory6,9).

Fig. 1A,B. On which side of the inserted forceps do the venous openings into the inferior vena cava (IVC) lie? An experiment in which a metal wire was used instead of forceps was performed to determine the answer. Dorsal view of the liver is shown, with A right and B left courses of the wire (the intermediate course is not shown). The right course of the wire lies between the openings of the inferior right hepatic vein (white arrow) and the caudate vein (black arrow), whereas in the left course, the wire crosses the opening of the caudate vein. Other small venous openings were also identied after this experiment: veins from the portal vein branch of hilar bifurcation origin (HB), veins of right portal vein origin (R), and veins of S7 and S8 origin (7 and 8). The caudal part of Spiegels lobe (asterisks) is supplied by the portal vein branch of HB origin. Stars, pocket-like spaces between the terminal portions of the right and middle hepatic vein (RHV and MHV); GB, gallbladder; LAL, left anatomical lobe; SP, Spiegels lobe

Results General observations Examination of the 56 livers revealed 63 thick caudate veins (over 3 mm in diameter) draining Spiegels lobe, and 26 IRHVs were identied in 25 of the 56 livers. The largest caudate vein was 9.0 mm in diameter (average,

T.J. Sato et al.: Anatomy for hanging maneuver without liver mobilization

57

the territory of the other short hepatic veins, although these veins were removed during the deep dissection. Distribution of venous openings into the IVC The distribution of the venous openings is shown in Fig. 3. The openings of the caudate vein, IRHV, and MRHV were located on either the left side or right side of the IVC, although some openings converged relatively to the left side. The openings of the caudate vein were concentrated in the middle one-third of the craniocaudal axis of the retrohepatic IVC, whereas the openings of the IRHVs were concentrated in the caudal one-third. The openings of the SHV-LC and SHV-RC, especially the former, tended to be dispersed. In some specimens, the SHV-LC even emptied into the IVC to the right of the SHV-RC. Even in the absence of any IRHVs (31 livers), there were no major differences from the other specimens (25 livers with IRHVs) in the distribution of the venous openings inside the IVC. Experiment using a metal wire instead of forceps We identied the veins located to the left and right of the path of the wire or along the wire when the wire was positioned along three routes that corresponded to possible paths of the forceps inserted blindly in the procedure of Belghiti et al.1 (Fig. 1). The results are summarized in Tables 13. The openings of the caudate veins were consistently located to the left of the right course, whereas the terminal portions of the IRHV and MRHV were also often located to the left of the wire (Table 1). Moreover, tracing of the right course revealed that the SHV-RC (right territory) as well as the SHV-LC (left caudate drainage territory) usually emptied into the IVC on the left side of the wire. By contrast, tracing of the left course (Table 3) showed that the IRHV and MRHV almost always lay on the right side of the wire, whereas the caudate vein was almost always on the left side. The SHV-LC, however, often drained into the IVC on the right side of the wires course.

Fig. 2A,B. Deep dissections to identify the drainage territory of the veins. A Dorso-caudal view and B dorsal view. A and B are different specimens. A We rst identied the courses of the caudate vein, inferior right hepatic vein (IRHV), and middle right hepatic vein. The inferior vena cava IVC and Spiegels lobe (SP) have been reected cranially. In these specimens, the thick caudate veins (stars) are clearly seen, whereas the other short hepatic veins (arrows) are numerous and thin. Note that the caudate veins run along the IVC along their terminal courses. B The next step was the dissection to identify the borders between S6, S7, S8, and the three territories of the caudate branches of the portal vein: the left portal territory (LC), the territory of the branch of hilar bifucation origin (HB), and the right portal territory (RC). The caudate vein (stars) has been cut and partly removed. This Spiegels lobe contains a notch (a white paper has been inserted in it). LigV, Ligamentum venosum; PT, portal trunk; large asterisk, pathological indent; small asterisk, dorsolateral paracaval branch of the posterior sectorial trunk

Discussion The implements used in the actual operations performed by Belghiti et al.1 and in our experiment were different: we used a thin metal wire instead of relatively thick forceps. When inserted from the caudal side of the liver toward the pocket-like space between the MHV and RHV, the width of the forceps is about 815 mm in the middle one-third and caudal one-third of the retrohepatic IVC. This seems to explain why not only all

5.1 mm), whereas the largest IRHV was 12.0 mm in diameter (average, 7.4 mm). During dissection from the dorsal and cranial aspects of the liver (Fig. 2) performed after the experiment described below, we checked to see whether the caudate vein originated from Spiegels lobe irrespective of whether or not a thin tributary from the paracaval portion joined the vein. We also identied

58

T.J. Sato et al.: Anatomy for hanging maneuver without liver mobilization

Fig. 3. Distribution of the openings of veins, with different drainage territories, into the IVC. The openings of the veins into the IVC in the 56 specimens are superimposed on a standardized schematic diagram of the IVC. This is a dorsal view after the dorsal wall of the IVC has been cut and reected. The caudate vein openings are concentrated in the middle one-third and on the left side of the retrohepatic IVC, whereas the IRHV joins the IVC in its caudal one-third and on the right side. There was no considerable difference between the distributions of the opening of the SHV-LC (veins from the left portal territory) and the SHV-RC (veins from the right portal territory)

Table 1. Right course of the metal wire on the dorsal aspect of the liver and comparison of the locations of the points of entry of the veins into the IVC Caudate vein Left side of the wire In the path of the wire Right side of the wire Total 63 veins (56 livers) 0 (0) 0 (0) 63 (56) SHV-LC 41 (27) 3 (2) 1 (1) 45 (30) SHV-RC 24 (13) 0 (0) 5 (3) 29 (16) IRHV 8 (8) 13 (13) 5 (4) 26 (25) MRHV 8 (7) 9 (8) 29 (24) 46 (39) SHV-8 3 (2) 1 (1) 0 (0) 4 (3)

Caudate vein, 1 or 2 thick veins over 3 mm in diameter at their junctions with the inferior vena cava (IVC) and originating from Spiegels lobe; SHV-LC, thin drainage vein, 13 mm in diameter, arising from the left half of the caudate lobe; SHV-RC, thin drainage vein, 13 mm in diameter, from the right half of the caudate lobe; IRHV, 1 or 2 thick veins over 5 mm in diameter at their junctions with the IVC and arising from S6; MRHV, drainage veins, over 5 mm in diameter, from S7 including the dl-paracaval branch territory; SHV-8, drainage vein(s), over 1 mm in diameter, from S8 including the PV8c territory

Table 2. Intermediate course of the metal wire and comparison of the locations of the points of entry of the veins into the IVC Caudate vein Left side of the wire In the path of the wire Right side of the wire Total 59 veins (52 livers) 4 (4) 0 (0) 63 (56) SHV-LC 28 (17) 11 (9) 6 (4) 45 (30) SHV-RC 10 (6) 9 (5) 10 (5) 29 (16) IRHV 2 (2) 4 (4) 20 (19) 26 (25) MRHV 2 (2) 3 (3) 41 (34) 46 (39) SHV-8 0 (0) 0 (0) 4 (3) 4 (3)

Abbreviations are same as those in the Table 1

Table 3. Left course of the wire and comparison of the locations of the points of entry of the veins into the IVC Caudate vein Left side of the wire In the path of the wire Right side of the wire Total 45 veins (40 livers) 16 (14) 2 (2) 63 (56) SHV-LC 13 (9) 8 (5) 24 (16) 45 (30) SHV-RC 0 (0) 8 (5) 21 (11) 29 (16) IRHV 0 (0) 1 (1) 25 (24) 26 (25) MRHV 0 (0) 1 (1) 45 (38) 46 (39) SHV-8 0 (0) 0 (0) 4 (3) 4 (3)

Abbreviations are same as those in the Table 1

T.J. Sato et al.: Anatomy for hanging maneuver without liver mobilization

59

veins along the wire (see Tables 13) but also some veins alongside the wire were damaged during surgery. When the forceps were inserted along the proper course, however, preservation of the caudate vein seemed possible because the openings of the caudate vein was usually localized to the left side of wire. However, the openings of both the IRHV and the caudate vein were sometimes found to the left of the wires course, possibly because of the long extrahepatic course of the IRHV. Moreover, the area of the potential space free of venous openings, even if such an area existed, was usually narrower, i.e., about 515 mm, than the width of the forceps (i.e., 815 mm: see above), making preservation of both of the caudate vein and the IRHV difcult. Therefore, some pretreatment, such as taping the IRHV before inserting the forceps, would seem to be effective if it is necessary to preserve the IRHV because it drains a large territory. Moreover, if care is taken to preserve the IRHV, it would seem that the MRHV would also be preserved on the right side of the forceps, because of its location further to the right of the openings than the IRHV. Consequently, to preserve the caudate vein during preparation of extended left-lobe liver grafts by the hanging maneuver without mobilizing the liver, we recommend the right course in our study as the path for insertion of the forceps, with attention being paid to the question of preservation of the IRHV. Takayama et al.4 preserved the caudate vein during living-related liver transplantation, using a left-lobe liver graft that included the caudate lobe, whereas Miyagawa et al.13 sacriced the caudate vein, and the good results obtained by this group cast doubt on role of the caudate vein in drainage; moreover, according to their results, thin short hepatic veins, such as those denoted as the SHV-LC and SHV-RC in the present study, were not responsible for venous drainage of the caudate lobe. Therefore, the problem concerning these thin veins seems to be limited to bleeding during and after insertion of the forceps. Nevertheless, Belghiti et al.1 reported no problem with such bleeding along the IVC. Our recent study14 suggested that the high content of smooth muscle tissue in the adventitia of the IVC stimulated rapid hemostasis in small hepatic veins. If the caudate vein is successfully preserved during insertion of the forceps, the next question is how to nd and reconstruct the vein in the left-lobe liver graft. It may be difcult to nd the caudate vein during anterior dissection without taping the vein and without mobilization of the liver. In contrast to the taping of the IRHV, taping of the caudate vein appeared to be difcult without liver mobilization. However, we think that surgeons will be able to locate the caudate vein in the nal step of anterior dissection if they separate the caudate lobe from the IVC, because the terminal course of the caudate vein runs along the IVC (Fig. 2).

In addition to the caudate vein arising from Spiegels lobe (usually, from a large part of the left half of the caudate lobe), we found many other veins arising from the left half, i.e., the SHV-LC, as has been reported by Gadzijev et al.15 and Filipponi et al.;16 however, our study revealed intermingling of the openings of the veins from the right half of the lobe (SHV-RC), rather than segregation of the distribution of their openings. Moreover, the openings of the SHV-RC also intermingled with those of the IRHV, MRHV, and S8 veins (Fig. 3). Therefore, the short hepatic veins do not indicate either the right margin of the caudate lobe or the border between the right and left halves. Our failure to identify these borders of the caudate lobe according to the venous openings was in clear contrast to the results of previous studies based on portal ramication (Ungvary,17 Yamamoto,18 Kogure et al.19). Our ndings suggest that, regardless of whether the maneuver of Belghiti et al.1 is used or the usual procedure with mobilization is performed, not only the SHV-LC but also the SHV-RC (and the S8 veins as well) are likely to be cut or damaged along the IVC during the preparation of extended left liver grafts. In the present study, the hilar bifurcation territory is included into the left caudate lobe (see Materials and methods). However, in conventional method for preparation of left-lobe liver grafts, it might be difcult to reconstruct the caudate branch of hilar bifurcation origin. Moreover, according to our very recent study, whether the hilar bifurcation branch is present or absent is committed to how the right portal territory extends into Spiegels lobe (now submitted). Thus recognition of the hilar bifurcation territory would provide a new impact on left-lobe liver grafts as well as conventional lobectomy. Acknowledgments. We are grateful to the following professors and their laboratory staff for the use of their specimens: (in alphabetical order) Professor H. Abe, Akita University School of Medicine; Professor Y. Dodo, Tohoku University Graduate School of Medicine; Professor Y. Fukui, Tokushima University School of Medicine; Professor M. Kikuchi, Tohoku University School of Dentistry; Professor S. Kitamura, Tokushima University School of Dentistry; Professor J. Matsumura, Kyorin University School of Medicine; Professor T. Sato, Tokyo Medical and Dental University Graduate School; and Professor T. Yajima, Health Science University of Hokkaido School of Dentistry.

References
1. Belghiti J, Guevara OA, Noun R, Saldinger PF, Kianmanesh R (2001) Liver hanging maneuver: a safe approach to right hepa-

60

T.J. Sato et al.: Anatomy for hanging maneuver without liver mobilization
tectomy without liver mobilization. J Am Coll Surg 193:109 111 Chang RW, Shan-Quan S, Yen WW (1989) An applied anatomical study of the ostia venae hepaticae and the retrohepatic segment of the inferior vena cava. J Anat 164:4147 Camargo AM, Teixeira GG, Ortale JR (1996) Anatomy of the ostia venae hepaticae and retrohepatic segment of the inferior vena cava. J Anat 188:5964 Takayama T, Makuuchi M, Kubota K, Sano K, Harihara Y, Kawarasaki H (2000) Living-related transplantation of left liver plus caudate lobe. J Am Coll Surg 190:635638 Makuuchi M, Hasegawa H, Yamazaki S, Bandai Y, Watanabe G, Ito T (1983) The inferior right hepatic vein: ultrasonic demonstration. Radiol 148:213217 Kumon M (1985) Anatomy of the caudate lobe with special reference to the portal vein and bile duct (in Japanese with English abstract). Kanzo (Acta Hepatol Jpn) 26:11931199 Yamane T, Mori K, Sakamoto K, Ikei S, Akagi M (1988) Intrahepatic ramication of the portal vein in the right and caudate lobes of the liver. Acta Anat 133:162172 Kogure K, Kuwano H, Fujimaki N, Makuuchi M (2000) Relation among portal segmentation, proper hepatic veins, and external notch of the caudate lobe in the human liver. Ann Surg 231:223 228 Kitagawa S, Murakami G, Hata F, Hirata K (2000) Conguration of the right portion of the caudate lobe with special reference to identication of its right margin. Clin Anat 13:321340 Kwon DH, Murakami G, Wang HJ, Chung MS, Hata F, Hirata K (2001) Ventral margin of the paracaval portion of human caudate lobe. J Hepatobiliary Pancreat Surg 8:148153 11. Couinaud C (1981) Controlled hepatectomies and exposure of the intrahepatic bile ducts. Anatomical and technical study. Couinaud, Paris, pp 2021 12. Ishiyama S, Yamauchi H (2000) Anatomy of the caudate lobe using corrosion liver casts (in Japanese). Geka (Surgery) 62:426 433 13. Miyagawa S, Hashikura Y, Miwa S, Ikegami T, Urata K, Terada M, Kubota T, Nakata T, Kawasaki S (1998) Concomitant caudate lobe resection as option for donor hepatectomy in adult living related liver transplantation. Transplantation 66:661663 14. Kanamura T, Murakami G, Hirai I, Hata F, Sato TJ, Kumon M, Nakajima Y (2001) High dorsal drainage routes of Spiegels lobe. J Hepatobiliary Pancreat Surg 8:549556 15. Gadzijev EM, Ravnik D, Stanisavljevic D, Trotovsek B (1997) Venous drainage of the dorsal sector of the liver: differences between segments I and IX. Surg Radiol Anat 19:7983 16. Filipponi F, Romagnoli P, Mosca F, Couinaud C (2000) The dorsal sector of human liver: embryological, anatomical and clinical relevance. Hepato-Gastroenterology 47:17261731 17. Ungvary G (1977) Functional morphology of the hepatic vascular system. Akademiai Kiado, Budapest, pp 5964 18. Yamamoto K (1987) Hepatic segment: a study from a vascular architectural point of view (in Japanese with English abstract). Kanzo (Acta Hepatol Jpn) 28:949961 19. Kogure K, Kuwano H, Fujimaki N, Makuuchi M (2000) Relation among portal segmentation, proper hepatic vein, and external notch of the caudate lobe in the human liver. Ann Surg 231:223 228

2.

3.

4.

5.

6.

7.

8.

9.

10.