EJSO 33 (2007) 729e734

www.ejso.com

Outcome following repeat liver resection for colorectal liver metastases

H. Nishio 1, Z.Z.R. Hamady, H.Z. Malik, S. Fenwick, K. Rajendra Prasad,
G.J. Toogood, J.P.A. Lodge*
HPB and Transplant Unit, St Jamess University Hospital, Beckett Street, Leeds LS9 7TF, UK
Accepted 4 July 2006
Available online 26 January 2007

Abstract
Aim: Our aim was to determine independent predictors of survival after second liver resection and to confirm whether the type of first resection influences survival after repeat resection.
Methods: Fifty-four patients who underwent a second liver resection for colorectal liver metastases were analyzed. To find independent
predictors of survival, possible prognostic factors regarding the primary tumor, and the first and second resections were used in the Cox
regression analysis.
Results: There were three postoperative deaths within 90 days of surgery. The 3- and 5-year overall survival rates were 53% and 46%, respectively. The size of the tumor (>50 mm) ( p 0.005), serum carcinoembryonic antigen level (>30 mg/L) ( p 0.002), and the presence
of a positive surgical margin at the second resection ( p 0.006) were independent predictors of poor survival following the second resection. The type of first resection was not associated with survival but was associated with the ability to achieve a histological negative surgical margin at the second liver resection ( p 0.01).
Conclusion: Three independent predictors of survival were identified. Major initial liver resection was associated with a reduced ability to
achieve surgical clearance at the second resection. For colorectal liver metastases, major resection should only be performed if a negative
margin cannot be achieved by minor resection.
2006 Elsevier Ltd. All rights reserved.
Keywords: Hepatectomy; Liver metastases; Colorectal neoplasm

Introduction
Surgical resection colorectal liver metastases (CRLM)
can offer 5-year survival rates of 20e40%,1e4 whereas untreated patients have a poor median survival of less than
12 months5,6 and chemotherapy can only extend this
median survival to 12e18 months.7,8 However, 60% of
patients undergoing hepatic resection for CRLM will develop recurrent disease.9e12 One-third of these patients
will have isolated recurrent liver metastases and will be
candidates for repeat resection.9e13
Many reports have been published analyzing the results
of repeat resection for CRLM, with most series having analyzed less than 50 patients.14e26 What is evident from the
literature is the fact that the type of initial resection
* Corresponding author. Tel.: 44 113 206 4890; fax: 44 113 244 8182.
E-mail address: peterlodge@aol.com (J.P.A. Lodge).
1
Visiting surgeon from Nagoya University on a Japanese National
Scholarship to work in the HPB and Transplant Unit in Leeds.
0748-7983/$ - see front matter 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.ejso.2006.07.005

(anatomical vs. non-anatomical resection) does not influence survival.19,23,24,26 Therefore, it has been advocated
that as long as the surgical margin can be secured, extensive
resections should be avoided. The aim of this study is to analyze the results of repeat resection for CRLM within our
tertiary referral HPB center.
Patients and methods
Patients
From January 1993 to May 2004, 540 patients underwent liver resection for CRLM at the HPB and Transplant
Unit, St. Jamess University Hospital, UK. Fifty-four of the
540 patients underwent a repeat liver resection for recurrent
liver metastases. Patients with localized liver recurrence, in
the absence of unresectable extra-hepatic disease (as defined by CT and MRI scans) and who were considered fit
for further surgery, underwent repeat resection. Resection

730

H. Nishio et al. / EJSO 33 (2007) 729e734

was performed using the Cavi-Pulse Ultrasonic Surgical

Aspirator (CUSA, Model 200T, Valley Lab., Colorado,
USA), with Pringle maneuver applied only in selected cases
to minimize blood loss.
According to our unit protocol, all patients were offered
systemic adjuvant chemotherapy with 5-fluorouracil and folinic acid unless they had undergone chemotherapy adjuvant
to bowel resection or primary hepatic resection within
12 months of repeat resection. Neoadjuvant chemotherapy
was offered to patients who were judged not to be initially
unresectable in order to down size the disease. In such cases
Oxaliplatin or Irinotecan has been used in recent years.

Results
Patient characteristics and primary tumor details
There were 34 male patients and 20 females within this
cohort. Mean age was 61 years (range, 24e79 years) at the
time of repeat liver resection. The primary tumor was colonic cancer in 31 patients and rectal cancer in 23 patients.
Thirty-four patients had synchronous liver metastases
(Dukes D), 13 patients were Dukes C and 8 were either
Dukes A or B.
Operative procedures

Surveillance after surgery

Following initial postoperative review, all patients were
examined in the outpatient clinic at 3, 6, 12, 18 and 24
months and annually thereafter. The minimum follow-up
period from repeat resection was 25 months (IQR
12e40). At each clinic visit blood was drawn for liver function tests and carcinoembryonic antigen (CEA). Additionally, at 3, 6, 12, and 18 months, and 2, 3, 4, 5, 7 and 10
years computed tomography (CT) of chest, abdomen and
pelvis was performed. Liver-magnetic resonance imaging
(MRI) and positron emission tomography (FDGePET)
were used to define suspicious CT lesions or in cases of
negative CT with rising tumor markers.
Data collection and statistical analyses
Patients data were collected from a prospective computerized database, patients medical records. Follow-up data
were obtained from patients reviews, medical records and
the patients registry. Major hepatic resection was defined
as the resection of three or more segments of the Couinaud
classification. Extra-hepatic disease was defined as infiltration of contiguous structure(s), presence of distant metastases or local recurrence of the primary colorectal tumor.
Mortality within 30 days of surgical resection was recorded.
Postoperative complications (morbidity) were recorded
within the hospital stay period. Continuous and categorical
variables were compared using the unpaired t-test and the
Chi-squared test accordingly, where appropriate. Longterm survival was calculated using the KaplaneMeier
method and compared by the log-rank test. A multivariate
analysis for survival was performed by the Cox regression
(forward stepwise model) including variables that would
be expected to influence outcome. Variables entered into
the equation were as follows: age, lymph nodes status of
the primary colorectal cancer, extent of first hepatic resection (minor/major), disease free interval between first and
second resections, CEA level at second resection, number
and size of liver metastases at first and second hepatic resections and resection margin involvement after second liver
resection. A 2-tail p-value less than 0.05 was considered
to indicate statistical significance.

Surgical procedures undertaken in the first and repeat resections are summarized in Table 1. Major resection was
performed in 31 patients at the time of initial hepatic resection. A further 8 patients underwent major resection at the
repeat resection. Thus, after the repeat resection in all, 39
patients had undergone a major resection.
Adjuvant and neoadjuvant chemotherapies
At the first hepatic resection, 11 patients had undergone
neoadjuvant chemotherapy and a further 30 patients received adjuvant chemotherapy. A further patient underwent
both adjuvant and neoadjuvant chemotherapies. At the time
of repeat hepatic resection, 5 patients received neoadjuvant
chemotherapy with a further 12 patients receiving adjuvant
chemotherapy. Furthermore, another patient underwent
both adjuvant and neoadjuvant chemotherapies.
Extra-hepatic disease
At the time of repeat resection, 13 (24%) patients had
extra-hepatic disease and 5 of these had lung metastases.
Three of the 5 patients with lung metastases underwent
Table 1
Types of surgical resection
Initial
resection

Repeat
resection

Major resection
Right trisectionectomy (segments 4, 5,
6, 7, 8  1)  PR
Left trisectionectomy (segments 2, 3, 4,
5, 8  1)  PR
Right hemi-hepatectomy (segments 5, 6,
7, 8  1)  PR
Left hemi-hepatectomy (segments 2, 3,
4  1)  PR

31
13

8
4

14

Minor resection
One sectionectomy  PR
Partial resection

23
8
15

46
3
43

Major resection is defined as the resection of three or more segments of the

Couinaud classification.
PR: partial resection of the liver or liver remnant.

H. Nishio et al. / EJSO 33 (2007) 729e734

curative lung surgery and were considered disease free after

the lung surgery. The other 8 patients, who had intraperitoneal extra-hepatic tumors, were considered as disease free
after the second hepatectomy because all of the macroscopic extra-hepatic disease had been removed.

731

The actuarial 3- and 5-year survival rates following repeat resection were 53% and 46%, respectively (Fig. 1).
The median survival was 50.3 months (95% Confidence
Interval 21e80).
Predictors of poor outcome

Morbidity and mortality after repeat resection

Ten patients had postoperative complications after the
repeat resection including hemorrhage (n 5); bile leak
(n 4); transient liver failure (defined as acutely deranged
liver functions followed by full recovery) (n 1); sepsis
(n 1) and/or other major complications (n 6). Three patients died during postoperative course. One was a 73-yearold man who underwent right trisectionectomy with partial
resection of the small intestine and the diaphragm after first
partial resection of segments 5 and 6 died following pulmonary embolus. Another 79-year-old man who underwent
right anterior sectionectomy after first right posterior sectionectomy died of methicillin resistant Staphylococcus aureus sepsis. The third patient was a 59-year-old male
undergoing partial resection of segment 7 after first left trisectionectomy, caudate lobectomy and partial resection of
right posterior section died of a secondary hemorrhage.
Outcome after repeat resection
Thirty-eight patients developed recurrence of whom 12
underwent further surgery. Four patients underwent resection for lung metastases and 2 underwent resection for local
recurrence of their primary colorectal cancer. There were 6
patients who underwent a third resection for recurrent liver
metastases of whom 2 underwent a forth resection for further liver metastases. At the end of follow-up 32 patients
were alive, 13 of them were disease free.
1.0
.9
.8

Survival Rate

.7
.6
.5
.4
.3
.2
.1
0.0
0

12

24

36

48

60

72

84

96

Months
Number at risk
0

12

24

36

48

60

72

84

Median survival

54

38

26

14

50.3

Figure 1. Survival curve for 54 patients following repeat resection.

(Survival was calculated as time from the repeat resection.)

Univariate analysis of prognostic factors influence on

survival is shown in Table 2. Multivariate analysis, using
prognostic factors with a possible impact on survival including the type of the first hepatectomy, found that the
size of the tumor at the second hepatectomy (50 mm vs.
>50 mm) (HR 5.2, 95%CI 1.6e16.6, p 0.005), serum
CEA level before the second hepatectomy (30 mg/L vs.
>30 mg/L) (HR 6.8, 95%CI 2e23, p 0.002) and surgical resection margin at the second hepatectomy (negative
vs. positive) (HR 5, 95%CI 1.5e15.9, p 0.006) were
Table 2
Univariate analysis predictors for poor survival after repeat resection
p-Value

Variable

Number

Median
survival
(month)

Age
<60
>60

25
29

64
50

0.610

Disease free interval H1eH2

<12 Months
>12 Months

24
30

64
50

0.914

Type of resection at H1
Minor
Major

23
31

35.1
50.3

0.478

Number of metastasis at H1
Single
Multiple

19
35

64.1
48.1

0.760

Size of metastases at H1
<50 mm
>50 mm

34
20

64
50.3

0.366

Surgical margin at H1
Negative
Positive

25
29

50.1
50.3

0.857

Number of metastasis at H2
Single
Multiple

40
14

50.3
32.9

0.531

Size of metastases at H2
<50 mm
>50 mm

44
10

64.1
27.6

0.031

Preoperative CEA at H2
<30 mg/L
>30 mg/L

42
12

64.2
24

<0.001

Surgical margin at H2
Negative
Positive

25
29

64.1
27.6

0.017

Blood transfusion at H2
No
Yes

42
12

50.3
64.1

0.864

H1: first resection; H2: repeat resection.

H. Nishio et al. / EJSO 33 (2007) 729e734

732

independent predictors of survival. Fig. 2 represents the actuarial survival of patients stratified according to the above
predictors of poorer outcome.
Impact of type of initial resection
Univariate analysis revealed that a major initial resection
resulted in a positive surgical resection margin during the
repeat resection in 19 (61%) patients compared to 6
(26%) patients who had undergone an initial minor resection ( p 0.01) (Table 3).
Discussion
Repeat resection for recurrent CRLM is technically demanding because of the development of severe perihepatic
adhesions, fragile liver parenchyma due to chemotherapy
and altered anatomy in the remnant liver due to regeneration.
Despite these limitations, surgeons have advocated repeat
resection as the treatment of choice even in patients with
resectable extra-hepatic disease because of a demonstrated
acceptable morbidity and mortality rate, and long-term survival.19,23 Following this aggressive surgical strategy, we
obtained a median survival of 50 months with morbidity
and mortality rates comparable with other reported series.17,19,25,27 Recently, Elias et al.28 have reported the usefulness of percutaneous radiofrequency ablation as an
.. A

1.0

0.36

0.04

___ B

.9

0.01

.8

Survival Rate

.7
.6
.5
.4
.3
.2
.1
0.0
0

12

24

36

48

60

72

84

96

Months
Number at risk
Months

12

24

36

48

60

72

18

14

24

20

14

84

Median survival
72

69
25

Figure 2. Survival curves according to the number of negative independent

predictors for survival. Size of the largest tumor more than 50 mm, serum
CEA level more than 30 mg/L and positive surgical margin at repeat resection are negative independent predictors of survival after the repeat hepatic
resection. (A) Patients without any negative independent predictors. (B)
Patients with one of the 3 negative independent predictors. (C) Patients
with more than one of the 3 negative independent predictors.

Table 3
Impact of the initial resection upon repeat resection
Initial resection
Minor

p-Value

Variable

Major

Disease free interval between

H1 and H2 (mean  SD, year)
Number of liver tumor-H2
(solitary/multiple)
Largest size of liver tumor-H2
(50 mm/>50 mm)
Serum CEA level-H2
(30 mg/L or >30 mg/L)
Extra-hepatic disease-H2
(present/absent)
Additional procedure-H2 (yes/no)
Intra-operative blood
transfusion-H2 (yes/no)
Surgical margin-H2
(positive/negative)
Postoperative complication
(present/absent)
Hospital death (yes/no)

1.3 (1.01) 1.28 (0.94)

0.903

21/10

11/12

0.14

26/5

21/2

0.68

2/24

4/14

0.2

5/26

8/15

0.2

5/26
24/6

7/16
17/6

0.3
0.6

19/12

6/17

0.01

5/25

5/18

0.7

1/30

2/21

0.6

H2: repeat resection; SD: standard deviation.

alternative to repeat hepatectomy for recurrent liver metastases and stressed that repeat hepatectomy is indicated only
when percutaneous radiofrequency ablation is contraindicated or fails because it is less invasive. However, in their series, one patient died from complications and the follow-up
time was only 14 months, providing a 2-year survival rate of
55% compared to 75% in our series. Long-term follow-up is
mandatory to determine the correct strategy for the management of CRLM.
Despite many reports regarding repeat resection for
CRLM, most have included less than 50 patients14e20 and
few authors23e25 have carried out a multivariate analysis
with more than 50 patients in order to find independent predictors for survival (Table 4). In our series, positive surgical
margin, size of metastases and a high serum CEA level
were independent predictors of poor survival after repeat
resection.
In our study, the type of initial resection, namely major
resection was associated with a positive resection margin at
the time of repeat resection and this independently affected
survival. We and others have emphasized that the achievement of a clear surgical margin is more important than the
type of resection for initial hepatectomy.29,30 Our results
suggest that once a major resection has been performed,
this limits the surgical options that may be available.
Some surgeons like to perform a major hepatic resection
for small colorectal metastases to obtain more than 1-cm
surgical margin. However, the 1-cm surgical margin theory
remains controversial and major hepatic resection limits the
choice of the next procedure. We have shown that 1-mm
surgical margin clearance is satisfactory to achieve acceptable survival in CRLM. Recently, Adam et al.31 have analyzed 60 third resections and found that although the type
of first, second or third hepatic resection was not associated

H. Nishio et al. / EJSO 33 (2007) 729e734

733

Table 4
Previous reports of repeat resection with more than 50 resections providing independent predictors (H2: repeat resection) compared to current study
Author, year

Number
of centers

Adam et al., 1997

Yamamoto et al., 1999

Number
of patients

Morbidity
(%)

Mortality
(%)

5-Year
survival (%)

Median survival
(months)

Independent predictors of poor outcome

64

20

41

46

75

11

31

31

Petrowsky et al., 2002

126

28

1.6

34

37

Current study

54

19

46

50

Curative nature of H2 ( p 0.008)

Interval < 1 year between H1 and H2
( p 0.009)
Number of tumor at H2 > 3
Positive regional LN metastases at H2
Extra-hepatic disease at H2
Residual tumor at H2
Number of tumor at H2 > 1 ( p 0.01)
Size of tumor at H2 > 50 mm ( p 0.04)
Curative nature of H2 ( p 0.006)
Size of tumor at H2 > 50 mm ( p 0.005)
Serum CEA at H2 > 30 mg/L ( p 0.002)

with poorer survival following the third hepatectomy,

a non-curative pattern of third liver resection is one of independent prognostic factors contributing to reduced survival.
Thus, it seems sensible to suggest that if a negative surgical
margin can be obtained, a minor resection should be chosen
at least as the first procedure as major hepatectomy may affect the ability to achieve a negative surgical margin at subsequent resection.
In conclusion, we have shown that a proportion of patients undergoing repeat resection for CRLM can have
a long-term survival. Furthermore, three independent prognostic factors predicting poorer outcome following repeat
hepatic resection have been identified, including one technical factor: the resection margin. These factors should be
taken into account when counseling patients for hepatic resection, and when stratifying patients for trials of neoadjuvant chemotherapies. Re-resection after initial major
resection was associated with a reduced ability to achieve
surgical clearance at repeat resection. Thus we suggest that
for CRLM major hepatectomy should only be performed if
a negative margin cannot be achieved by minor resection.

8.

9.

10.
11.

12.

13.
14.

15.

16.

References
1. Scheele J, Stangl R, Altendorf-Hofmann A, Paul M. Resection of colorectal liver metastases. World J Surg 1995;19:5971.
2. Nordlinger B, Guiguet M, Vaillant JC, et al. Surgical resection of colorectal carcinoma metastases to the liver. A prognostic scoring system
to improve case selection, based on 1568 patients. Association Francaise de Chirurgie. Cancer 1996;77:125462.
3. Fong Y, Cohen AM, Fortner JG, et al. Liver resection for colorectal
metastases. J Clin Oncol 1997;15:93846.
4. Elias D, Cavalcanti A, Sabourin JC. Results of 136 curative hepatectomies with a safety margin of less than 10 mm for colorectal metastases. J Surg Oncol 1998;69:8893.
5. Jaffe BM, Donegan WL, Watson F, Spratt Jr JS. Factors influencing
survival in patients with untreated hepatic metastases. Surg Gynecol
Obstet 1968;127:111.
6. Bengmark S, Hafstrom L. The natural history of primary and secondary malignant tumours of the liver. Cancer 1969;23:198202.
7. Chang AE, Schneider PD, Sugarbaker PH, Simpson C, Culnane M,
Steinberg SM. A prospective randomized trial of regional versus

17.

18.
19.

20.
21.

22.
23.
24.

systemic continuous 5-fluorodeoxyuridine chemotherapy in the

treatment of colorectal liver metastases. Ann Surg 1987;206:68593.
Cunningham D, Pyrhonen S, James RD, et al. Randomized trial of irinotecan plus supportive care versus supportive care alone after fluorouracil failure for patients with metastatic colorectal cancer. Lancet
1998;352:14138.
Bozzetti F, Doci R, Bignami P, Morabito A, Gennari L. Patterns of
failure following surgical resection of colorectal cancer liver metastases. Ann Surg 1987;205:26470.
Ekberg H, Tramberg KG, Anderson R, et al. Patterns of recurrence in
liver resection for colorectal secondaries. World J Surg 1987;11:5417.
Lange JF, Leese T, Castaing D, Bismuth H. Repeated hepatectomy for
recurrent malignant tumors of the liver. Surg Gynecol Obstet 1989;
169:11926.
Stone MD, Cady B, Jenkins RL, McDermott WV, Steele Jr GD. Surgical therapy for recurrent liver metastases from colorectal cancer.
Arch Surg 1990;125:71822.
Griffith KD, Sugarbaker PH, Chang AE, et al. Repeated hepatic resections for colorectal metastases. Br J Surg 1990;77:2303.
Fong Y, Blumgart LH, Cohen AM, Fortner J, Brennan MF. Repeat hepatic resections for metastatic colorectal cancer. Ann Surg 1994;220:
65762.
Pinson CW, Wright JK, Chapman WC, Garrard CL, Blair TK,
Sawyers JL. Repeat hepatic surgery for colorectal cancer metastases
to the liver. Ann Surg 1996;223:76576.
Tuttle TM, Curley SA, Roh MS. Repeat hepatic resection as effective
treatment for recurrent colorectal liver metastases. Ann Surg Oncol
1997;4:12730.
Chu QD, Vezeridis MP, Avradopoulos KA, Wanebo HJ. Repeat
hepatic resection for recurrent colorectal cancer. World J Surg 1997;
21:2926.
Kin T, Nakajima Y, Kanehiro H, et al. Repeat hepatectomy for recurrent colorectal metastases. World J Surg 1998;22:108791.
Muratore A, Polastri R, Bouzari H, Vergara V, Ferrero A,
Capussotti L. Repeat hepatectomy for colorectal liver metastases:
a worthwhile operation? J Surg Oncol 2001;76:12732.
Suzuki S, Sagaguci T, Yokoi Y, et al. Impact of repeat hepatectomy on
recurrent colorectal liver metastases. Surgery 2001;129:4218.
Nordlinger B, Vaillant JC, Guiguet M, et al. Survival benefit of repeat
liver resections for recurrent colorectal metastases: 143 cases. J Clin
Oncol 1994;12:14916.
Fernandez-Trigo V, Shamsa F, Sugarbaker PH. Repeat liver resection
from colorectal metastasis. Surgery 1995;117:296304.
Adam R, Bismuth H, Castaing D, et al. Repeat hepatectomy for colorectal liver metastases. Ann Surg 1997;225:5162.
Yamamoto J, Kosuge T, Shimada K, Yamasaki S, Moriya Y,
Sugihara K. Repeat liver resection for recurrent colorectal liver metastases. Am J Surg 1999;178:27581.

734

H. Nishio et al. / EJSO 33 (2007) 729e734

25. Petrowsky H, Gonen M, Jarnagin WR, et al. Second liver resection are
safe and effective treatment for recurrent hepatic metastases from colorectal cancer. A bi-institutional analysis. Ann Surg 2002;235:86371.
26. August DA, Sugarbaker PH, Ottow RT, Gianola FJ, Schneider PD. Hepatic resection of colorectal metastases. Influence of clinical factors
and adjuvant intraperitoneal 5FU via Tenckhoff catheter. Ann Surg
1985;201:2108.
27. Zacharias T, Jaeck D, Oussoultzoglou E, Bachellier P, Weber J-C. First
and repeat resection of colorectal liver metastases in elderly patients.
Ann Surg 2004;240(5):85865.
28. Elias D, De Baere T, Smayra T, Ouellet JF, Roche A, Lasser P. Percutaneous radiofrequency thermoablation an alternative to surgery for

treatment of liver tumour recurrence after hepatectomy. Br J Surg

2002;89(6):7526.
29. Tanaka K, Shimada H, Ohta M, et al. Procedures of choice for resection of primary and recurrent liver metastases from colorectal cancer.
World J Surg 2004;28:4827.
30. Hamady ZZR, Cameron IC, Wyatt J, Prasad RK, Toogood GJ,
Lodge JPA. Resection margin in patients undergoing hepatectomy
for colorectal liver metastasis: a critical appraisal of the 1 cm rule.
Eur J Surg Oncol 2006;32(5):55763.
31. Adam R, Azoulay D, Tanaka K, Castaing D, Bismuth H. Liver resection for colorectal metastases: the third hepatectomy. Ann Surg 2003;
238(6):87183.