REVIEW ARTICLE

Int J Clin Oncol (2005) 10:97–102 © The Japan Society of Clinical Oncology 2005DOI 10.1007/s10147-004-0481-6

Junji Yamamoto · Akio Saiura · Rintaro KogaMakoto Seki · Masashi Ueno · Masatoshi OyaKaoru Azekura · Yasuyuki Seto · Shigekazu OhyamaSatoshi Fukunaga · Toshiharu YamaguchiNorihiro Kokudo · Masatoshi Makuuchi · Tetsuichiro Muto

Surgical treatment for metastatic malignancies. Nonanatomical resection ofliver metastasis: indications and outcomes

Received: January 31, 2005

Key words Colorectal cancer · Liver metastasis · Surgery ·Survival · Surgical margin

Introduction

Surgical resection represents the only radical treatment forliver metastases from colorectal cancer, and has been inves-tigated at many centers.1–31 Prospective comparative studiesare needed to more accurately verify the advantages ofresection. However, now that surgical resection is consid-ered to have a certain survival benefit and other treatmentmodalities show a low likelihood of cure, the studies thatare actually being conducted should be those comparingsurgical resection with non-surgical treatment in patientsfor whom surgical resection is of some limited use.

Surgeons express differing opinions regarding the selec-tion of resection procedures (anatomical hepatectomy vsnonanatomical limited hepatectomy) and the extent of sur-gical margins. The present report discusses the findings of aliterature review focusing on these contentious issues.

Indications for surgical resection of colorectalliver metastases

Surgical resection is indicated in patients with metastaticliver cancer if the preoperative and intraoperative diag-noses indicate that the tumor mass can be safely and com-pletely resected. The safety of liver resection depends on:

(1) functional hepatic reserve after resection; and (2) thedegree of difficulty of the surgical procedure. These factorsare, in turn, determined by the location of the tumor(s)within the liver and the extent of liver involvement. Hepaticreserve is evaluated by estimating anticipated residual livervolume and function based on liver function tests, and im-aging information obtained using computed tomographyor other methods. Hepatic reserve does not often limit theindications for resection, because liver metastasis usuallyoccurs in normal livers. In patients who require extensiveresection of uninvolved sections of liver (e.g., extendedright hepatectomy plus partial resection of the left liver andtrisegmentectomy plus partial resection of the remainingsegments), the risk of postoperative liver failure due tomassive parenchymal loss can be reduced by performingportal vein embolization of the area to be resected, ratherthan one-stage hepatectomy.32–34

The degree of difficulty of liver resection is determinedby the relationship between the tumor and the hilar hepato-portal region, inferior vena cava, and major hepatic veins.As both liver resection itself and vascular reconstructioncan now be performed safely, few patients are considered tohave “nonresectable” disease as a result of tumor character-istics (tumor diameter, distribution, and location), otherthan the number of metastases. Liver resection combinedwith resection and reconstruction of the hepatic veins andinferior vena cava should also be performed whenever pos-sible if complete resection can be achieved.35,36 Even if thetumor has infiltrated the confluence of the hepatic veins andinferior vena cava, complete resection can be achieved byresecting and reconstructing the main hepatic vein.37

Previously, the conditions that were considered to pre-clude resection in patients with colorectal liver metastasiscomprised: (1) liver metastases that cannot be completelyresected; (2) regional hepatic lymph nodes positive for me-tastasis; and (3) extrahepatic remote metastases.38 How-ever, now that liver resection can be performed extremelysafely, indications for surgical resection are expanding insome areas. For example, relatively good outcomes havebeen reported even in patients with pulmonary metastasis ifliver resection is also performed, provided that the pulmo-

J. Yamamoto (*) · A. Saiura · R. Koga · M. Seki · M. Ueno ·M. Oya · K. Azekura · Y. Seto · S. Ohyama · S. Fukunaga ·T. Yamaguchi · T. MutoDepartment of Surgery, Cancer Institute Hospital, 3-10-6 Ariake,Koto-ku, Tokyo 135-8550, JapanTel. �81-3-3520-0111; Fax �81-3-3570-0343e-mail: jyamamoto@jfcr.or.jp

N. Kokudo · M. MakuuchiHepatobiliary Pancreatic Surgery Division, Graduate School ofMedicine, University of Tokyo, Tokyo, Japan

98

nary metastases are completely resectable.39–41 Cliniciansmust recognize that failing to operate on patients with com-pletely resectable localized metastases overlooks thosepatients who may be curable.

Outcomes after surgical resection of colorectal livermetastases

Surgical outcomes in the main studies of resection forcolorectal liver metastases are shown in Table 1. Operativemortality in recent studies was 0–3%, and the 3-, 5-, and10-year survival rates after surgery were 31%–57%, 16%–51%, and 19%–28%, respectively. Overseas studies alsoincluded a recent study that reported a 5-year survival rateexceeding 50%.31 Establishment of patient selection criteriaand improvements in resection techniques were thoughtto have contributed to this result. The breakdown of dataon patients who have undergone resection is important.In most European and American studies, 50% or more ofpatients had tumors with a maximum diameter of more than5cm. In contrast, only around 30% of patients in Japanesestudies had tumors of this size. Conversely, around 40% ofpatients in Japanese studies displayed multiple bilobardisease, compared to only about 10%–30% in overseasstudies. Patients with four or more metastases comprisedless than 10% of subjects in European and American stud-ies, contrasting sharply with the 23% in Japanese studies.These differences suggest variations in patient selectioncriteria based on resection techniques.

Factors influencing surgical outcome

Direct comparison of factors affecting surgical outcomeafter resection of colorectal liver metastases is difficult, asthe factors included in analysis differ from study to study(Table 2). Little divergence of opinion is seen with regard tothe following as prognostic factors: (1) stage of the primarydisease; (2) number of metastases (�4 vs �4); (3) status ofsurgical margins (positive vs negative); (4) metastasis tohepatic regional lymph nodes; (5) extrahepatic disease; and(6) satellite nodules. The following have been reported asspecial prognostic factors based on pathological examina-tions of resected specimens: (1) tumor pseudocapsule; (2)macroscopic invasion of the bile duct; (3) intrahepaticvascular invasion;42 (4) lymphatic duct invasion;43 and (5)islands of entrapped liver cells in metastases.44 In contrast,evaluations of the following differ depending on the study:(1) time of diagnosis of liver metastasis (synchronous vsmetachronous); (2) intrahepatic distribution of metastases(unilobar vs bilobar); (3) resection procedure (nonanato-mical resection vs anatomical resection); and (4) extent ofsurgical margin (�10mm vs �10mm). Obtaining uniformresults would, presumably, be difficult because the criteriaused to select patients for resection differ from study tostudy, and numerous factors are involved in the prognosis

of metastatic liver cancer, due to a complex pathology thatincludes the primary tumor.

Liver resection procedures

Hepatocellular carcinoma is well-known to display a highaffinity for the portal vein, and shows a transportal patternof spread.45 If colorectal liver metastases show markedsecondary intrahepatic spread, the resection procedureused should be in accordance with the pattern of spread.Our review showed that, in colorectal liver metastases, boththe number of metastases and invasion of the portal vein bymetastases were indicators of poor postoperative outcome,but that the two indicators were unrelated. Multiple livermetastases thus appear to occur not as a result of intrahe-patic metastasis by invasion of the portal vein, but, rather,as a result of the formation of multiple metastases fromthe outset.29 Accordingly, surgical treatment of liver me-tastasis does not require close adherence to a systematicapproach based on anatomy of the portal tract. Instead,limited hepatectomy is generally used, so that completeresection can be achieved in as many patients as pos-sible.29,42,46 Using intraoperative ultrasonography, tumorsare resected so as to maximize the sparing of uninvolvedliver parenchyma and achieve tumor-free surgical margins.If patients display multiple metastases in one hemiliverand hemihepatectomy or segmentectomy appears consider-ably easier than individually resecting tumors by limitedresection, hemihepatectomy or segmentectomy will beused. In overseas studies, major anatomical resections suchas hemihepatectomy represent the basic procedure used, asshown in Table 1.

Surgical margins

When considering surgical margins, a distinction needs tobe made between “whether tumor involvement is present atthe surgical margin” and “whether the surgical margin is10mm or more, or less than 10mm”. Postoperative outcomeis known to be appreciably worse in patients with tumorinvolvement of the surgical margin, as has already beenshown. The involvement of the surgical margin by the tu-mor would increase the likelihood of microscopic residualdisease; thus, why this should be avoided wherever possibleis easy to understand. However, the effects of the extent ofsurgical margins on postoperative outcome are not as clear.Some studies have found that postoperative outcomesdiffer depending on whether the surgical margin was 10mmor more, or less than 10mm,10,47,48 whereas others haveidentified no such differences.29,49 Our review showed thatpostoperative survival rates tended to be lower in patientswith surgical margins of 10mm or more than in patientswith surgical margins of less than 10mm on univariateanalysis. However, the extent of surgical margins does notrepresent an independent prognostic factor on multivariateanalysis. Extent of surgical margins has been shown to be

99

Tab

le 1

.St

udie

s of

out

com

e af

ter

surg

ical

res

ecti

on o

f co

lore

ctal

live

r m

etas

tase

s

Aut

hors

(ye

ar)

Num

ber

of p

atie

nts

who

3-Y

ear

5-Y

ear

50%

Sur

viva

lR

ecur

renc

eH

epat

icM

etas

tase

sSo

litar

y: M

ulti

ple

Surg

ical

or

unde

rwen

t re

sect

ion

surv

ival

rat

esu

rviv

al r

ate

peri

odra

te (

%)

recu

rren

ce�

4un

iloba

r: M

ulti

ple

in-h

ospi

tal

(num

ber

of p

atie

nts

who

(%)

(%)

rate

(%

)(%

)bi

loba

rm

orta

lity

unde

rwen

t ra

dica

l res

ecti

on)

(%)

Ads

on e

t al

.6 (19

80)

3441

5.9

Mor

row

et

al.7 (

1982

)a64

3420

Raj

pal e

t al

.8 (19

82)

3431

.3 M

onth

s11

.8F

ortn

er e

t al

.9 (19

84)

6557

7A

dson

et

al.10

(19

84)

141

252

Aug

ust

et a

l.11 (

1985

)33

53 (

4-ye

ar38

Mon

ths

0su

rviv

al r

ate)

Pet

relli

et

al.12

(19

85)

3622

Mon

ths

14 (

5/36

)B

ozze

tti e

t al

.13 (

1986

)45

3662

36–

26:1

4:5

NA

Gen

nari

et

al.14

(19

86)

4853

30 M

onth

s46

31–

26:1

5:7

2.1

But

ler

et a

l.15 (

1986

)62

5034

6132

–42

:13

:79.

7Iw

atsu

ki e

t al

.16 (

1986

)60

(60

)53

4548

3012

–0

Ekb

erg

et a

l.17 (

1986

)72

3016

22 M

onth

s78

6519

34:1

8:1

65.

6N

ordl

inge

r et

al.18

(19

87)

8040

.524

.964

43–

44:1

9:1

75

Hol

m e

t al

.20 (

1989

)35

3174

660

11:1

6:8

0Sc

heel

e et

al.21

(19

90)

226

(183

)38

6143

105

:40

:28

5 (1

2/22

6)D

oci e

t al

.22 (

1991

)10

7 (1

00)

3069

41–

58:2

8:1

45

Ros

en e

t al

.23 (

1992

)28

047

252.

8 Y

ears

––

––

3.6

Nak

amur

a et

al.24

(19

92)

3145

6148

19–

3Su

giha

ra e

t al

.25 (

1993

)15

9 (1

09)

57.2

(C

urat

ive

47.9

(C

urat

ive

6032

51:1

6:4

01

rese

ctio

n)re

sect

ion)

Gay

owsk

i et

al.26

(19

94)

204

4332

7272

–91

:33

:80

0Ja

tzko

et

al.27

(19

95)

6629

.673

38–

39:1

3:1

44.

5Y

amam

oto

et a

l.29 (

1999

)96

6151

––

–23

39:1

7:4

00

Min

agaw

a et

al.30

(20

00)

235

5138

3.1

Yea

rs77

–23

110

:36

:99

0C

hoti

et

al.31

(20

02)

226

5740

46 M

onth

s80

–9

141

:32

:53

1aIn

clud

es p

rim

ary

tum

ors

othe

r th

an c

olor

ecta

l can

cer

NA

, not

ava

ilabl

e

100

related to number of metastases.29 When removing tumorsthat require multiple resections, surgical margins will moreoften have to be less than 10mm to ensure sufficient re-sidual liver volume. While some studies have reported ahigher frequency of micrometastases in the liver paren-chyma within 10mm of the tumor margin,28 these are in theminority, and satellite configurations such as that seen withfungating intrahepatic cholangiocarcinoma are rarely seenwith colorectal liver metastases. Kokudo et al.50 conducted adetailed study of micrometastases around resected livermetastases, using K-ras and p53 mutations as genetic mark-ers. Micrometastases in the liver parenchyma were seen in2% of specimens, while metastases via the portal tract wereseen in 14.3%. All micrometastases were within close prox-imity (�5mm) of the tumor. They proposed a surgicalmargin of 2mm as the minimum requirement for colorectalliver metastases, and stated that margins of less than2mm carried a risk of margin-related recurrence ofapproximately 6%.

Specifying the extent of surgical margins as a require-ment for resection is undesirable, as this limits the numberof patients for whom resection is indicated. For example,the requirement that a surgical margin of 10mm or moreis the necessary condition for curative resection wouldexclude numerous patients with multiple bilobar livermetastases. When choosing suitable procedures for liverresection, a degree of flexibility is useful, and surgical mar-gins of 10mm should be used only as a guiding principle.This allows resection to be safely performed in a large num-ber of patients. In a relatively recent paper, by Cady et al.,46

ability to ensure surgical margins of 10mm was considereda condition for resection, but this criterion would markedlylower the resection rate. A comparison of studies conductedin Europe and the United States with those conducted inJapan shows distinct differences in the indications forpatients with multiple bipolar disease and patients withmultiple disease with four or more metastases (Table 1).The extent of surgical margins that can be ensured shouldnot be made a condition for liver resection, as the extent ofsurgical margins should be determined as a balance be-tween the number of metastases, locations of metastaseswithin the liver, and the resection procedure.

Conclusions

The present article has discussed resection of colorectalliver metastases based on a review of the literature. Al-though the response rate for nonsurgical treatment ofcolorectal liver metastases has increased, such treatment isfrequently not curative. To reiterate, the goal of treatmentshould be to completely remove tumors by resection when-ever possible. The treatment of first choice for resectablecolorectal liver metastases is surgical resection, and surgicalmargins should be sufficient to safely achieve completeresection.

Acknowledgments This work was supported in part by Grants-in-Aidfor Basic Science Research from the Ministry of Education, Culture,Sports, Science, and Technology.

Table 2. Factors affecting surgical outcome after resection of colorectal liver metastases

Sex Age Preoperative Synchronous Disease-free Site Stage of Size of(years) CEA value vs interval of primary primary metastatic

metachronous tumor tumor tumor(s)

Positivity rate (%) 20 10 33 38 0 10 50 20Ekberg17 (1986) � � � � � �Registry of Hepatic (�70 vs � � � �

Metastases19 (1988) �70Holm20 (1989) � � � � �Scheele21 (1990) � � � � � �Sugihara25 (1993) � � � � �Gayouski26 (1994) � � � � � �Jatzko27 (1995) � � � � � �Nordlinger18 (1987) � � � � �Beckerts (1997) � � � �

Jaeck (1997) � � � �Jamison (1997) � �Jenkins (1997) � � � �

Rees45 (1997) � � �Yasui (1997)Cady46 (1998) � � � �Elias (1998)Yamamoto29 (1999) � � � �Minagawa30 (2000) � � � � � � � �Choti31 (2002) � � � � �

�, Significant; �, not significant; �, not determined significant

101

References

1. Brunschwig A (1963) Hepatic lobectomy for metastatic cancer.Cancer 16:277–282

2. Woodington GF, Waugh JM (1963) Results of resection of meta-static tumors of the liver. Am J Surg 105:24–29

3. Flanagan L, Foster JH (1967) Hepatic resection for metastaticcancer. Am J Surg 113:551–557

4. Attiyeh FF, Wanebo HJ, Stearns MW (1978) Hepatic resection formetastasis from colorectal cancer. Dis Colon Rectum 21:160–162

5. Wanebo HJ, Semoglou C, Attiyeh F, et al. (1978) Surgical manage-ment of patients with primary operable colorectal cancer andsynchronous liver metastases. Am J Surg 135:81–85

6. Adson MA, Van Heerden JA (1980) Major hepatic resection formetastatic colorectal cancer. Ann Surg 191:576–583

7. Morrow CE, Grage TB, Sutherland DER, et al. (1982) Hepaticresection for secondary neoplasms. Surgery 92:610–614

8. Rajpal S, Dasmahapatra KS, Ledesma E, et al. (1982) Extensiveresections of isolated metastases from carcinoma of the colon andrectum. Surg Gynecol Obstet 155:813–816

9. Fortner JG, Silva JS, Golbey RB, et al. (1984) Multivariate analysisof a personal series of 247 consecutive patients with liver me-tastases from colorectal cancer. I. Treatment by hepatic resection.Ann Surg 199:306–315

10. Adson MA, Van Heerden JA, Adson MH, et al. (1984) Resectionof hepatic metastases from colorectal cancer. Arch Surg 119:647–651

11. August DA, Sugarbaker PH, Ottow RT, et al. (1985) Hepaticresection of colorectal metastases. Influence of clinical factors andadjuvant intraperitoneal 5-fluorouracil via Tenckhoff catheter onsurvival. Ann Surg 201:210–218

12. Petrelli NJ, Nambisan RN, Herrera L, et al. (1985) Hepatic resec-tion for isolated metastasis from colorectal carcinoma. Am J Surg149:205–209

13. Bozzetti F, Bignami P, Morabito A, et al. (1986) Patterns offailure following surgical resection of colorectal cancer liver

metastases. Rationale for multimodal approach. Ann Surg 205:264–270

14. Gennari L, Doci R, Bozzetti F, et al. (1986) Surgical treatmentof hepatic metastases from colorectal cancer. Ann Surg 203:49–54

15. Butler J, Attiyeh FF, Daly JM (1986) Hepatic resection formetastases of the colon and rectum. Surg Gynecol Obstet 162:109–113

16. Iwatsuki S, Esquivel CO, Gordon RD, et al. (1986) Liver resectionfor metastatic colorectal cancer. Surgery 100:804–809

17. Ekberg H, Tranberg KG, Andersson R, et al. (1986) Determinantsof survival in liver resection for colorectal secondaries. Br J Surg73:727–731

18. Nordlinger B, Parc R, Delva E, et al. (1987) Hepatic resection forcolorectal liver metastases. Influence on survival of preoperativefactors and surgery for recurrence in 80 patients. Ann Surg205:256–263

19. Registry of Hepatic Metastases (1988) Resection of the liver forcolorectal carcinoma metastases: a multi-institutional study of indi-cation for resection. Surgery 103:278–288

20. Holm A, Bradley E, Aldrete JS (1989) Hepatic resection ofmetastasis from colorectal carcinoma. Morbidity, mortality andpattern of recurrence. Ann Surg 209:428–434

21. Scheele J, Stangl R, Altendorf-Hoffmann A (1990) Hepatic me-tastases from colorectal carcinoma: impact of surgical resection onthe natural history. Br J Surg 77:1241–1246

22. Doci R, Gennari L, Bignami P, et al. (1991) One hundred patientswith hepatic metastases from colorectal cancer treated by resec-tion: analysis of prognostic determinants. Br J Surg 78:797–801

23. Rosen CB, Nagorney DM, Taswell HF, et al. (1992) Perioperativeblood transfusion and determinants of survival after liver resectionfor metastatic colorectal carcinoma. Ann Surg 216:493–505

24. Nakamura S, Yokoi Y, Suzuki S, et al. (1992) Results of extensivesurgery for liver metastases in colorectal carcinoma. Br J Surg79:35–38

25. Sugihara K, Hojo K, Moriya Y, et al. (1993) Pattern of recurrenceafter hepatic resection for colorectal metastases. Br J Surg80:1032–1035

Number of Number of Unilobar/ Satellite Resection Surgical Surgical Extrahepatic Hepatic Perioperativemetastatic metastatic Bilobar metastases procedure margin margin metastasis lymph bloodtumor(s) tumor(s) distribution Positive/ �10 mm node transfusionSolitary/ Multiple; Negative or �10mm metastasisMultiple �4 vs �4

53 75 7 100 36 100 50 55 100 0� � � � � � � �� � � � � � �

� � � �� � � � � � � �� � � �� � � � � � � � �� � � � �� � � �

� Incompleteresection

� � � �� � � � �� � Incomplete �

resection� � �� � � �� � �

� � �� � � � �� � � � � � �� � � � �

102

26. Gayowski TJ, Iwatsuki S, Madariage JR, et al. (1994) Experiencein hepatic resection for metastatic colorectal cancer: analysis ofclinical and pathologic risk factors. Surgery 116:703–711

27. Jatzko GR, Losborg PH, Stettner HM, et al. (1995) Hepatic resec-tion for metastases from colorectal carcinoma – a survival analysis.Eur J Cancer 31A:41–46

28. Shirabe K, Takenaka K, Gion T, et al. (1997) Analysis of prognos-tic risk factors in hepatic resection for metastatic colorectal carci-noma with special reference to the surgical margin. Br J Surg84:1077–1080

29. Yamamoto J, Shimada K, Kosuge T, et al. (1999) Factorsinfluencing suvival of patients undergoing hepatectomy forcolorectal metastasis. Br J Surg 86:332–337

30. Minagawa M, Makuuchi M, Torzilli G, et al. (2000) Extension ofthe frontiers of surgical indications in the treatment of liver me-tastases from colorectal cancer: long-term results. Ann Surg 231:487–499

31. Choti MA, Sitzmann JV, Tiburi MF, et al. (2002) Trends in long-term survival following liver resection for hepatic colorectal me-tastases. Ann Surg 235:759–766

32. Makuuchi M, Thai BL, Takayasu K, et al. (1990) Preoperativeportal embolization to increase safety of major hepatectomy forhilar bile duct carcinoma: a preliminary report. Surgery 107:521–527

33. Kawasaki S, Makuuchi M, Kakazu T, et al. (1994) Resection formultiple metastatic liver tumors after portal embolization. Surgery115:674–677

34. Nagino M, Nimura Y, Kamiya J, et al. (1995) Changes in hepaticlobe volume in biliary tract cancer patients after right portal veinembolization. Hepatology 21:434–439

35. Takayama T, Nakatsuka T, Yamamoto J, et al. (1996) Re-reconstruction of a single remnant hepatic vein. Br J Surg 83:762–763

36. Sakamoto Y, Yamamoto J, Kosuge T, et al. (2004) Extended lefthepatectomy by severing all major hepatic veins with reconstruc-tion of the right hepatic vein. Surg Today 34:482–484

37. Hughes K, Scheele J, Sugarbaker PH (1989) Surgery for colorectalcancer metastatic to the liver. Optimizing the results of treatment.Surg Clin North Am 69:339–359

38. Murata S, Moriya Y, Akasu T, et al. (1998) Resection of bothhepatic and pulmonary metastases in patients with colorectal carci-noma. Cancer 83:1086–1093

39. Robinson BJ, Rice TW, Strong SA, et al. (1999) Is resection ofpulmonary and hepatic metastases warranted in patients withcolorectal cancer? J Thorac Cardiovasc Surg 117:66–75; discussion75–76

40. Lehnert T, Knaebel HP, Duck M, et al. (1999) Sequential hepaticand pulmonary resections for metastatic colorectal cancer. Br JSurg 86:241–243

41. Yamamoto J, Sugihara K, Kosuge T, et al. (1995) Pathologic sup-port for limited hepatectomy in the treatment of liver metastasesfrom colorectal cancer. Ann Surg 221:74–78

42. Sasaki A, Aramaki M, Kawano K, et al. (2002) Prognosticsignificance of intrahepatic lymphatic invasion in patients withhepatic resection due to metastases from colorectal carcinoma.Cancer 95:105–111

43. Koike M, Yasui K, Torii A, et al. (2000) Prognostic significance ofentrapped liver cells in hepatic metastases from colorectal cancer.Ann Surg 232:653–657

44. Kokudo N, Tada K, Seki M, et al. (2001) Anatomical major resec-tion versus nonanatomical limited resection for liver metastasesfrom colorectal carcinoma. Am J Surg 181:153–159

45. Rees M, Plant G, Bygrave S (1997) Late results justify resection formultiple hepatic metastases from colorectal cancer. Br J Surg 84:1136–1140

46. Cady B, Jenkins RL, Steele GD Jr, et al. (1998) Surgical margin inhepatic resection for colorectal metastasis: a critical and improv-able determinant of outcome. Ann Surg 227:566–571

47. Sano K, Makuuchi M, Miki K, et al. (2002) Evaluation of hepaticvenous congestion: proposed indication criteria for hepatic veinreconstruction. Ann Surg 236:241–247

48. Maema A, Imamura H, Takayama T, et al. (2002) Impaired vol-ume regeneration of split livers with partial venous disruption: alatent problem in partial liver transplantation. Transplantation73:765–769

49. Hughes KS, Simon R, Songhorabodi S, et al. (1986) Resection ofthe liver for colorectal carcinoma metastases: a multi-institutionalstudy of patterns of recurrence. Surgery 100:278–284

50. Kokudo N, Miki Y, Sugai S, et al. (2002) Genetic and histologicalassessment of surgical margins in resected liver metastases fromcolorectal carcinoma: minimum surgical margins for successfulresection. Arch Surg 137:833–840