Hepatectomy for liver metastasis of colorectal cancer
- First Online:
- Cite this article as:
- Xu, J., Wei, Y., Zhong, Y. et al. Int J Colorectal Dis (2009) 24: 419. doi:10.1007/s00384-008-0619-5
- 123 Views
The objective of this study was to investigate whether hepatic resection (HR) can increase the survival of liver metastasis of colorectal cancer (CRC).
Materials and methods
CRC patients (n = 669) with liver metastasis treated at the Zhongshan Hospital, Fudan University from 1/2000 to 7/2007 were included in the study to investigate the relationship between HR and cancer survival.
CRC patients (n = 669) with liver metastases who had primary tumor resection were grouped in synchronous liver metastasis (SLM; 56.7%, n = 379) and metachronous liver metastasis (MLM) groups (43.3%, n = 290). Hepatic resection rates were lower (32.5%, n = 123) in the SLM than the MLM group (44.8%, n = 130, P < 0.05). The 30-day mortality rate in the MLM (2.3%) was significantly lower than SLM (2.4%) groups. The 5-year survival rates (36.6%) was same compared to SLM group (33.1%, P > 0.05). One-, 2-, and 3-year survival of stages I and II operation cases were 92.5% vs 86.5%, 0.7% vs 58.0%, and 42.1% vs 44.9% (P > 0.05) in the SLM group, respectively. Recurrence after first hepatic resection associated with a 2.23-fold increased risk of death (P < 0.01). Incision margins larger than 1 cm and HR for recurrence associated with 34% and 27% (P < 0.05) decreased death risk.
Hepatic resection could help the survival of liver metastasis of colorectal cancer, and stage I surgery is safe for this disease.
KeywordsColorectal neoplasmsLiverNeoplasm metastasisHepatectomy
Colorectal cancer is the fifth largest cancer in China. Its incidence is rising, especially in Shanghai, where it has moved to the second position with 5,507 new cases annually . Liver metastases is found in 25% of the patients when diagnosed. Untreated hepatic colorectal metastasis has a poor prognosis and is associated with a median survival of 6.9 months . Several retrospective studies have shown that liver resection prolongs survival. Fong et al.  reported 5- and 10-year survival rates of 37% and 22%, respectively, in a 13-year study of 1,001 liver resections for metastatic colorectal cancer. Xu et al.  reported 5-year survival rates of 29.4% in 133 resected patients compared with 0% in 49 judged resectable patients but not operated.
Hepatic resection has been associated with complication rates of 20% to 50% and perioperative mortality rates of 2% to 4% . A majority of patients who undergo hepatic resection develop recurrent diseases . But there is a large deviation between higher and lower volume hospitals , especially in developing countries such as China.
Zhongshan hospital is one of the hepatic centers in China, with more than 1,700 cases of hepatic operations, and has started surgical therapy for liver metastasis of colorectal cancer (LMCC) since 2000, and more than 250 hepatic resections have been performed. Perioperative complications, mortality, and survival analysis will be presented in this manuscript to convey some therapy points in China.
Materials and methods
Patients and data sources
Colorectal cancer patients (n = 669) with liver metastasis treated at the Zhongshan Hospital, Fudan University, between January 2000 and July 2007 were included in the study. Patients with any stage colorectal cancer at presentation who did not undergo primary tumor resection were excluded because of the paucity of hepatic resections in this group. Patients were divided into two groups. The first was metachronous liver metastasis (MLM) group and included patients with localized or regional disease (stages I through III) and who underwent colorectal resection and later developed hepatic metastases. The second group, or the synchronous liver metastasis (SLM) group, included patients with distant-stage disease (stage IV) at presentation who underwent primary tumor resection or patients who had a diagnosis of hepatic metastases within 6 months of their cancer diagnosis. Both groups was subdivided further into two comparison groups based on whether patients underwent liver resection or not.
Patient-level demographic and clinical variables identified from the case reports included age, sex, anatomic site, stage of disease at diagnosis (MLM or SLM), and level of serum carcinoembryonic antigen (CEA) and CA199.
Patients with hepatic metastases were identified from case reports, including partial lobectomy, trisegmentectomy, and total left or right lobectomy. The control of hepatic veins outflow before parenchymal division was nearly always obtained in lobectomy or extended resections. The liver parenchyma was divided to crush the liver tissue to expose bile ducts and blood vessels, which were then tied. Intermittent inflow vascular occlusion (the Pringle maneuver)  was applied at 5- to 10-min intervals, released briefly, and then reapplied as necessary. Pringle time was recorded as the total cumulative Pringle time applied during parenchymal transection.
In patients with hepatic resection, analysis of complications was limited to those that occurred during hospitalization for the first hepatic resection, including: postoperative hemorrhage, posthemorrhagic anemia, wound dehiscence, liver abscess, peritonitis, ileus, gastrointestinal hemorrhage, biliary fistula, intestinal fistula, stomach or duodenal fistula, and postoperative infection.
Perioperative mortality was defined as death within 30 days from the date of hepatic resection.
Categorical variables were compared by using the Pearson chi-square test or the Fisher exact test, where appropriate. P values <0.05 were considered significant. Data were analyzed with SPSS software (version 10.1 for Windows; SPSS Inc., Chicago, IL, USA).
Survival was calculated as the number of months from liver metastasis diagnosis to death and was censored at 5 years or at the latest date, since longer survival may not necessarily reflect death from colorectal cancer in this cohort of patients. Survivor functions were estimated by using the Kaplan–Meier method. The log-rank test was used to compare survival curves.
Variables that were associated on univariate analysis with significant differences in 5-year survival were entered into a multivariate Cox proportional hazards model. If the variables were no longer statistically significant in the multivariate model based on the partial likelihood ratio test, then they were removed in a stepwise fashion from the model unless we believed that they were meaningful clinically. Continuous variables were then examined for linearity in the logit, categorized as appropriate, and the model was refit with the recoded variables.
Clinical characterization of patients
Baseline characteristics of the MLM and SLM groups categorized by hepatic resection status
SLM group, 379 cases
MLM group, 290 cases
No HR, 256 cases
HR, 130 cases
No HR, 160 cases
Mean age ± SD, y
58.0 ± 13.6
58.7 ± 12.0
57.8 ± 10.7
58.7 ± 10.9
Ascending colon and cecum
LM time ± SD (months)
21.0 ± 19.6
265.5 ± 645.7
224.0 ± 410.1
65.1 ± 160.9
93.7 ± 206.0
792.2 ± 2185.7
1150.4 ± 2560.4
171.4 ± 354.9
495.7 ± 1552.9
Clinical features of hepatic resection cases
Hepatic resection rate %(cases)
P < 0.05
Number of liver lesions
P < 0.05
Size of lesions(cm)
P < 0.05
P < 0.05
P < 0.05
P < 0.05
P < 0.05
Pulmonary and cardiac
P > 0.05
P > 0.05
P > 0.05
P > 0.05
30-day mortality % (cases)
P > 0.05
Recurrence % (cases)
P < 0.05
Re-hepatic resection % (cases)
P > 0.05
Complications and 30-day mortality
First hepatic resection-associated complications were analyzed in patients. In total, 66 complications in 52 out of 253 (20.6%) patients were identified. The most common complication was wound dehiscence (11.1%, n = 28), followed by pleural effusion (7.1%, n = 18), pulmonary and cardiac complications (4.0%, n = 10), postoperative anemia (2.4%, n = 6), and anastomosis leakage (0.7% , n = 2). None of the patients had biliary fistula.
Of the 253 patients who underwent hepatic resection, 2.4% (n = 6) died within 30 days of surgery. The 30-day mortality rate in the MLM group was 2.3% (3/130), compared with 2.4% in SLM group (P > 0.05) patients presented initially with stage IV disease (Table 2).
The overall recurrence rate was 40.7% (103/253) including 57 cases (43.3%) in SLM and 45 (34.6%) cases in the MLM group (P < 0.05). In SLM group, 20 recurrent cases (35.1%) in the SLM and 35.6% (16/45) in the MLM group received re-hepatic resection (P > 0.05; Table 2).
Survival analysis of patients judged resectable
Among 669 cases, 312 patients were judged resectable, had indication for hepatic resection of liver metastasis, but only 253 cases received hepatic resection (HR). The indication included: (1) liver metastasis focus ≤4 or liver metastasis focus >4 but localized to semi-liver lobe; (2) the remnant liver volume ≥30%; (3) without other organ metastasis or peritoneum polar lymph node metastasis. The other 59 cases without liver metastasis were patients that refused to have the operation. The 1-, 3-, and 5-year survival rates were 84.5%, 51.3%, and 35.6% for HR patients and 56.7%, 10.1%, and 0% for no HR patients(P < 0.001; log-rank test, Fig. 1d), respectively, and the median survival was 37 and 15 months (P < 0.001; log-rank test).
Stages I and II operation in SLM group
In stage I operation group, 42 cases received colectomy (37 cases right colectomy and five cases transverse colectomy) and liver operation in one subcostal incision (one-incision subgroup). Another 29 cases received colectomy in left abdomen incision (five cases left colectomy, eight cases Sigmoid colectomy, nine cases low anterior resection, and seven cases abdominoperineal resection) and liver operation in another subcostal incision (two-incision subgroup).
Morbidity rate of II incision subgroup was 31.0% (9/29, including wound dehiscence (five cases), pleural effusion (one case), cardiac and pulmonary complications (two cases), and anastomosis leakage (one case)), which was similar to that of I incision subgroup (28.5% (12/42), including wound dehiscence (seven cases), pleural effusion (three cases), and cardiac/pulmonary complications (two cases)), P > 0.05. Thirty-day mortality of II incision subgroup (3.4%, 1/28) was a little higher than that of II incision subgroup (2.4%, 1/42), but there was no statistical significance (P > 0.05).
Cox proportional hazards model
Using separate multivariate Cox models, age at diagnosis, sex, primary tumor location, stage at presentation (SLM or MLM group), features of the liver metastasis (location, size, number), incision margin, postoperative therapy, recurrence after first hepatic resection, and the therapy of recurrence (HR or No HR) were examined to find risk factors for hepatic resection patients. The recurrence after the first hepatic resection was associated with a 2.23-fold increased risk of death (95% confidence interval (CI) 2.01 ∼ 2.53, P < 0.01; partial likelihood ratio test). Incision margin larger than 1 cm and HR for recurrence was associated with 34% (95% CI 0.48 ∼ 0.73, P < 0.05; partial likelihood ratio test) and 27% (95% CI 0.64 ∼ 0.80, P < 0.05; partial likelihood ratio test) decreased risk of death.
Hepatic colorectal metastasis is a suitable prototype disease treated by partial hepatectomy. It is now well-documented that partial hepatectomy can achieve long-term survival for selected patients with single or multiple colorectal liver metastases. Scheele et al.  reported 10- and 20-year survival rates of 23% and 18%, respectively in 434 liver resections performed between 1960 and 1992 for metastatic colorectal cancer. In contrast, for 983 patients thought to have unresectable tumor, the median survival was 6.9 months without any 5-year survivors . Sometimes surgery even effects a cure. D’Angelica et al.  reported that in 96 out of 456 resections, 5-year survivors subsequently had a 78% actuarial 10-year survival rate. No other treatment modality has demonstrated comparable results.
However, the results of present study show comparable results: disregarding the liver metastasis type, SLM or MLM, hepatic resection group can get 5-year survival to 36.6% and 33.1%, respectively. But compared to unresectable patients, hepatic resection patients have less and smaller liver metastasis focuses that can influence the survival. We also analyzed 312 judged resectable patients, including 59 patients who refused hepatic resection, with a median survival time of 15 months. However, the median survival time of hepatic resection patients reached to 37 months. Scheele et al.  reported a median survival of 30 months for resected patients compared with 14.2 months for resectable but not resected patients.
Surgical management of synchronous colorectal metastases is controversial and widely debated, and higher complication and mortality rates for patients with simultaneous resection of the colon and liver have been reported (stage I surgery) [9, 10]. Recently, large case reports indicated that with the advances in perioperative treatments and the surgical technology, safety for stage I surgery is the same as staged surgery. Chua et al.  reported that in 96 consecutive patients with synchronously recognized primary carcinoma and hepatic metastases who underwent stage I (n = 64) or II (n = 32) colonic and hepatic resections at the Mayo Clinic, perioperative morbidity was 53% and 41% (P = 0.25), with no operative mortality. Martin et al.  reported low perioperative mortality (2.2%) in stage I surgery patients (n = 134) comparable to 2.8% of 106 stage II surgery cases. Although stage I survival has been reported to be the same as stage I surgery, we found that 1- and 2-year survival could be increased by 6% (92.5% to 86.5%) and 12.7% (70.7% vs 58.0%) for stage I surgery but not the 3-year survival. More patients and longer follow-up times may be needed to confirm this observation. Recently, Peters et al.  analyzed tumor growth changes in synchronous liver metastases after the primary tumor resection and found some evidence for the choice of stage I surgery since the percentage of proliferating cells within the metastases showed a modest increase and a significant decrease in the percentage of apoptotic cells. Taken together, an increased net tumor growth of the metastases occurred after primary tumor resection.
Our results also show that incision margin larger than 1 cm was associated with 34% (95% CI 0.48 ∼ 0.73, P < 0.05; partial likelihood ratio test) decreased risk of death. Elias et al.  reported that, in a group of 75 LMCC patients with hepatectomy, 5-year survival rate was 42% with incision margin >1 cm or 25% with incision margin <1 cm. Pawlik et al.  found that 5-year survival of patients with incision margins of 1 ∼ 4 mm, 5 ∼ 9 mm, and ≥10 mm were 62.3%, 71.1%, and 63%, respectively. Patients with positive incision margin had a poor 5-year survival of 17.1%. This shows that incision margin should not be more than 1 cm, but if possible, it should not be less than 1 cm.
We also found that HR recurrence was associated with 27% decreased risk of death. Petrowsky et al.  reported that 1-, 3-, and 5-year survival rates were 86%, 51%, and 34% in 126 patients (American n = 96, 1986–2001; European n = 30, 1985–1999) who underwent repeat liver resection and 19 actual 5-year survivors to date. Adam et al.  reported overall 5-year survival of 32% and disease-free survival of 17% after the third resection in 60 patients with colorectal metastases in a 16-year study (1984–2000). The third hepatic resection was attempted in 68 of 115 of liver recurrences following a second hepatectomy (59%) and achieved in 88% of the cases (60 of 68). Survival rates compared favorably to patients with recurrences following a second hepatectomy who could not be operated (5% at 3 years) or who failed to be resected (15% at 2 years, P = 0.0001). This data suggest that if the recurrence is resectable, hepatectomy should be the first choice.
In the present retrospective study, more than 670 cases were included within 7 years, and our observations may be helpful for clinical treatment of liver metastasis of colorectal cancer.