Ann Surg Oncol DOI 10.1245/s10434-015-4460-0
ORIGINAL ARTICLE – COLORECTAL CANCER
Long-Term Outcomes of Conversion Hepatectomy for Initially Unresectable Colorectal Liver Metastases Yoshiaki Maeda, MD, PhD, Toshiki Shinohara, MD, Akihisa Nagatsu, MD, Noriaki Futakawa, MD, and Tomonori Hamada, MD, PhD Department of Surgery, Hokkaido Cancer Center, Sapporo, Japan
ABSTRACT Background. Chemotherapy, including molecular targeted agents, for metastatic colorectal cancer has greatly improved recently and offers an increased chance of conversion hepatectomy for patients with initially unresectable liver metastases. However, the long-term outcomes of conversion hepatectomy remain controversial. Methods. We retrospectively assessed a consecutive series of 210 patients with colorectal liver metastases to evaluate the long-term outcomes of patients who underwent conversion hepatectomy and to clarify the predictive factors related to the conversion rate. Results. Ninety-four cases were initially resectable and underwent primary hepatectomy. Of the 116 patients with initially unresectable liver metastases, 104 patients underwent chemotherapy (systemic or hepatic artery infusion). Twenty-four percent (11/46) of the initially unresectable liver-limited metastases that became resectable after chemotherapy were subsequently treated with conversion hepatectomy; however, there were no cases of conversion among the patients with extrahepatic metastases. The final resection rate of liver metastases was 50 % (105/210), including conversion hepatectomies. The predicted 5-year survival rate in the conversion hepatectomy group was 76 %. The conversion rate was significantly (P \ 0.05) higher in patients with liver-limited metastases (24 %), patients with no LN involvement (27 %), the hepatic arterial infusion chemotherapy group (33 %), patients treated with anti-EGFR agents (21 %), and patients with a complete or partial response (33 %).
Conclusions. Twenty-four percent of the patients with initially unresectable liver-limited metastases became resectable after chemotherapy, and the survival rate after conversion hepatectomy was not inferior to that of the primary hepatectomy cases. Chemotherapy regimens with high response rates are needed to achieve a higher conversion rate.
Ó Society of Surgical Oncology 2015
PATIENTS AND METHODS
Liver resection is the standard and only curative treatment for colorectal liver metastases, and the 5-year survival rate after complete resection for patients with colorectal liver metastases was reported to be 40–50 %.1–3 However, initial liver resection can be performed in only 20–30 % of the patients with liver metastases, and the 5-year survival rate of patients treated with chemotherapy alone is less than 5 %.4–7 Initially unresectable liver metastases may become resectable when marked tumor shrinkage is achieved after induction chemotherapy, and this is called conversion hepatectomy.8,9 Chemotherapy, including molecular targeted agents for metastatic colorectal cancer, has improved greatly within the past decade and offers an increased chance of conversion liver resection for patients with initially unresectable liver metastases. Recently, many reports have described downsizing by chemotherapy and rescue liver surgery for initially unresectable colorectal liver metastases; however, the long-term outcomes of conversion hepatectomy are still controversial.10,11 This study was designed to clarify the effectiveness of conversion hepatectomy for long-term survival in Japanese patients with colorectal liver metastases and to determine the factors associated with the conversion rate.
First Received: 19 October 2014 Y. Maeda, MD, PhD e-mail: [email protected]
A consecutive series of 210 patients with colorectal liver metastases in a single institute (Hokkaido Cancer Center)
Y. Maeda et al.
treated from 2000–2012 were included in this study. We retrospectively assessed the characteristics of the patients and their survival after the treatments. Primary hepatectomy was performed if technically possible. From 2009, neoadjuvant chemotherapy (6 cycles of FOLFOX) was administrated for primary resectable cases with three or more liver nodules. Resectability was decided based on the size of the remnant liver volume (expected 30 % functional liver remnant after the removal of all metastases), regardless of the number or the size of the liver metastases. Partial resection with a free margin of [1 cm was preferred, and anatomical resection was selected when needed. Even if extrahepatic metastases were present, hepatectomy was performed for the patients whose extrahepatic metastases were also resectable. All detectable lesions were resected to achieve R0 resection, and radiofrequency ablation also was used in some cases for small nodules (\10 mm) located in the deeper liver in some patients. Chemotherapy was administrated to the patients diagnosed with unresectable disease. As first-line chemotherapy, oxaliplatin-based and CPT-11 regimens were mainly used. Before 2008, hepatic artery infusion (HAI) chemotherapy was applied for selected patients. From 2007, an anti-VEGF agent was combined with the regimens of systemic chemotherapy. From 2008, antiEGFR agents were combined with the regimens of systemic chemotherapy for the patients who had wild-type KRAS status. Patients were followed by CT scanning and the resectability was assessed every 2 or 3 months by liver surgeons, medical oncologists, and radiologists. If curative
FIG. 1 Flow diagram of the 210 patients with colorectal liver metastases. Ninety-four patients were initially resectable at the time of diagnosis and underwent hepatectomy (primary resection group). Thirteen of them received neoadjuvant chemotherapy before hepatectomy. The other 116 patients were diagnosed to have unresectable liver metastases, and 104 of them received chemotherapy. Eleven patients converted to be resectable after chemotherapy and underwent hepatectomy (conversion group). Ninety-three patients remained unresectable even after chemotherapy (unresectable group). Twelve patients received best supportive care only
resection became possible, conversion hepatectomy was planed. Chemotherapy was continued until conversion hepatectomy could be performed or until the patient could no longer receive chemotherapy because of tumor progression or due to the adverse effects of the treatments. The overall survival curves from the date of the first treatment for liver metastases (hepatectomy or chemotherapy) were estimated using the Kaplan–Meier method and were analyzed using the log-rank test. To determine the independent relationships between preoperative variables and conversion hepatectomy, Fisher’s exact provability tests were performed. All statistical analyses were performed using the EZR statistical software program.12 P \ 0.05 was considered to be statistical significant. RESULTS A total of 210 patients with colorectal liver metastases were treated between 2000 and 2012. A flow diagram of the 210 patients with colorectal liver metastases is shown in Fig. 1. Ninety-four patients were initially resectable at the time of diagnosis and underwent hepatectomy (primary resection group). Thirteen of them received neoadjuvant chemotherapy before hepatectomy. Sixteen patients also had extrahepatic metastases (lung or others) and underwent resection of both the liver and extrahepatic metastases to achieve R0 resection. The other 116 patients were diagnosed to have unresectable lesions, 104 of whom received chemotherapy. The reasons for unresectability were as follows: liver limited factors in 46 cases (expected functional liver remnant \30 %), the presence of unresectable
210 patients with colorectal liver metastases
94 initially resectable
13 received NAC
116 initially unresectable
104 received chemotherapy
12 BSC only
94 received hepatectomy
11 received conversion hepatectomy
93 remained unresectable
Primaly resection group
Conversion group
Unresectable group
Conversion Hepatectomy for Unresectable Metastases
extrahepatic metastases in 21 cases, and both liver and extrahepatic factors in 37 cases. Eleven patients converted to be resectable after chemotherapy and underwent hepatectomy (conversion group). Ninety-three patients remained unresectable even after chemotherapy (unresectable group). Twelve patients received best supportive care only. The clinical characteristics of the initially resectable (n = 94) and unresectable (n = 116) patients are summarized in Table 1. A significantly higher incidence of extrahepatic metastases, shorter interval to liver metastases, higher number of liver metastatic nodules, larger size of liver metastases, and higher CEA level were observed in initially unresectable group compared to the resectable group. TABLE 1 Patient and tumor characteristics Initially resectable group (n = 94)
Initially unresectable group (n = 116)
P value
Age (years) \75
85
95
C75
9
21
0.11
Gender Female Male Extrahepatic disease
26
51
68
67
Present
16
61
None
78
55
0.022
Factors Predicting Conversion Hepatectomy after Chemotherapy \0.0001
Lymph node metastases N0
27
26
N(?)
63
80
4
10
Colon
59
91
Rectum
35
25
NX
0.42
Location of primary tumor 0.048
Interval to metastases Metachronous
27
5
Synchronous (\12 months)
67
111
81
36
C5 13 Size of largest liver metastasis
80
\0.0001
No. of metastases \5
\5 cm
83
59
C5 cm
11
37
0
20
\20
69
34
C20
25
82
Unknown
\0.0001
\0.0001
CEA level \0.0001
Italic values are used for names of gene or statistical figures (P value) CEA carcinoembryonic antigen
Among the 116 patients in the initially unresectable group, 86 received systemic chemotherapy, 18 received HAI chemotherapy, and 12 received best supportive care only. A weekly high-dose 5FU regimen was used for HAI chemotherapy, and CPT11 also was used as combination treatment in seven of these patients. The first-line protocols for systemic chemotherapy were as follows: CPT-11based regimens in 34 patients, oxaliplatin-based regimens in 40 patients and others (capecitabine, S1, 5FU/LV) in 11 patients. An anti-VEGF agent (bevacizumab) was added in 34 patients, and an anti-EGFR agent (cetuximab or panitumumab) was added in 19 patients with a wildtype KRAS status. The lesions in 11 patients (10.5 %) became resectable after chemotherapy and these subjects subsequently underwent R0 hepatectomy (conversion hepatectomy). The surgical procedures for hepatectomy in the conversion cases were as follows: right lobectomy in five cases (3 patients also received 2–4 wedge resections) and nonanatomical resection in six cases (1–12 wedge resections). One case of surgical mortality within 30 days was seen in primary resection group (1.1 % 1/94); however, no in-hospital death was observed in the conversion patients. The final rate of resection of the liver metastases was 50 % (105/210), including the cases of conversion hepatectomy.
The rate of conversion in the patients with liver-limited metastases was 24 % (11/46). In contrast, no cases of conversion were observed among the patients with extrahepatic metastases (0/58; P \ 0.0001; Table 2). Potential factors predicting conversion hepatectomy in the liver-limited cases are shown in Table 3. The conversion rate was significantly higher in the patients with a negative LN status of the primary lesion (50 %; P = 0.049), the patients treated with hepatic arterial infusion chemotherapy (46 %; P = 0.050), the patients treated with an anti-EGFR agent (44 %; P = 0.013), and the patients exhibiting a complete or partial response (58 %; P \ 0.0001). Long-Term Survival in Patients with Colorectal Liver Metastases Figure 2A shows the results of a Kaplan–Meier analysis of the overall survival in the primary hepatectomy group, the conversion group, the unresectable group, and the best supportive care group. The median follow-up period among the survivors at the last follow-up was 89 months in the conversion group and 56 months in the primary resection group. The overall survival of the conversion group
Y. Maeda et al.
was significantly higher than that of the unresectable group (P \ 0.0001). The predicted 5-year survival rate in the conversion hepatectomy group was 76 %, and five patients survived at least 5 years. Three of these patients survived more than 5 years tumor-free after their final treatments. No patient in the nonconversion group survived more than 5 years.
TABLE 2 Conversion rates according to the presence of extrahepatic metastases Extrahepatic disease
Conversion rate (%)
P value
Present
0
0/58
\0.0001
None (liver limited metastases)
24
11/46
Bold value indicates statistical significance (p \ 0.05)
TABLE 3 Factors predicting conversion hepatectomy in liver limited cases Conversion group (n = 11)
Unresected group (n = 35)
Odds ratio
95 % CI
P value
\75
11
27
–
–
0.17
C75
0
8 2.05
0.43–11.3
0.49
4.77
0.89–28.2
0.049
2.04
0.20–104.4
0.99
–
–
0.11
3.91
0.44–190.4
0.25
–
–
0.99
2.03
0.33–22.4
0.47
0.50
0.098–2.41
0.48
1.42
0.25–7.20
0.71
4.60
0.89–25.9
0.050
0.79
0.14–5.74
0.99
16.3
1.28–935.5
0.013
Age (years)
Gender Female
7
16
Male
4
19
Lymph node metastases N0
6
6
N(?)
5
25
NX
0
4
Histological differentiation of primary tumor Differentiated 10 Undifferentiated
29
1
6
Wild-type
7
8
Mutant
0
5
Unknown
4
22
10
25
1
10
KRAS status
Location of primary tumor Colon Rectum Interval to metastases Metachronous
1
0
10
35
C5
9
24
\5
2
11
5
22
6
13
\20
4
10
C20
7
25
Synchronous (\12 months) No. of metastases
Size of largest liver metastasis C5 cm \5 cm CEA level
Initial treatment for liver metastases HAI
6
7
Systemic chemotherapy
5
28
LOHP or CPT11 based
8
27
Others
3
8
Regimen
Anti-EGFR agenta Used
4
5
Not used
1
23
Conversion Hepatectomy for Unresectable Metastases TABLE 3 continued Conversion group (n = 11)
Unresected group (n = 35)
Odds ratio
95 % CI
P value
–
–
0.29
–
–
£0.0001
Anti-VEGF agenta Used
0
10
Not used
5
18
Response to chemotherapy CR or PR
11
8
SD or PD
0
27
Italic values are used for names of gene or statistical figures (P value). Bolditalic values are statistical significance (P \ 0.05) HAI hepatic artery infusion chemotherapy, LOHP oxialiplatin, CR complete response, PR partial response, SD stable disease, PD progressive disease a
Assessed in patients who underwent systemic chemotherapy only
A
B
Overall survival
1.0
Disease-free survival
0.1 Conversion group (n=11)
0.8
0.8
0.6
0.6 Primary resection group (n=94)
0.4
P
ORIGINAL ARTICLE – COLORECTAL CANCER
Long-Term Outcomes of Conversion Hepatectomy for Initially Unresectable Colorectal Liver Metastases Yoshiaki Maeda, MD, PhD, Toshiki Shinohara, MD, Akihisa Nagatsu, MD, Noriaki Futakawa, MD, and Tomonori Hamada, MD, PhD Department of Surgery, Hokkaido Cancer Center, Sapporo, Japan
ABSTRACT Background. Chemotherapy, including molecular targeted agents, for metastatic colorectal cancer has greatly improved recently and offers an increased chance of conversion hepatectomy for patients with initially unresectable liver metastases. However, the long-term outcomes of conversion hepatectomy remain controversial. Methods. We retrospectively assessed a consecutive series of 210 patients with colorectal liver metastases to evaluate the long-term outcomes of patients who underwent conversion hepatectomy and to clarify the predictive factors related to the conversion rate. Results. Ninety-four cases were initially resectable and underwent primary hepatectomy. Of the 116 patients with initially unresectable liver metastases, 104 patients underwent chemotherapy (systemic or hepatic artery infusion). Twenty-four percent (11/46) of the initially unresectable liver-limited metastases that became resectable after chemotherapy were subsequently treated with conversion hepatectomy; however, there were no cases of conversion among the patients with extrahepatic metastases. The final resection rate of liver metastases was 50 % (105/210), including conversion hepatectomies. The predicted 5-year survival rate in the conversion hepatectomy group was 76 %. The conversion rate was significantly (P \ 0.05) higher in patients with liver-limited metastases (24 %), patients with no LN involvement (27 %), the hepatic arterial infusion chemotherapy group (33 %), patients treated with anti-EGFR agents (21 %), and patients with a complete or partial response (33 %).
Conclusions. Twenty-four percent of the patients with initially unresectable liver-limited metastases became resectable after chemotherapy, and the survival rate after conversion hepatectomy was not inferior to that of the primary hepatectomy cases. Chemotherapy regimens with high response rates are needed to achieve a higher conversion rate.
Ó Society of Surgical Oncology 2015
PATIENTS AND METHODS
Liver resection is the standard and only curative treatment for colorectal liver metastases, and the 5-year survival rate after complete resection for patients with colorectal liver metastases was reported to be 40–50 %.1–3 However, initial liver resection can be performed in only 20–30 % of the patients with liver metastases, and the 5-year survival rate of patients treated with chemotherapy alone is less than 5 %.4–7 Initially unresectable liver metastases may become resectable when marked tumor shrinkage is achieved after induction chemotherapy, and this is called conversion hepatectomy.8,9 Chemotherapy, including molecular targeted agents for metastatic colorectal cancer, has improved greatly within the past decade and offers an increased chance of conversion liver resection for patients with initially unresectable liver metastases. Recently, many reports have described downsizing by chemotherapy and rescue liver surgery for initially unresectable colorectal liver metastases; however, the long-term outcomes of conversion hepatectomy are still controversial.10,11 This study was designed to clarify the effectiveness of conversion hepatectomy for long-term survival in Japanese patients with colorectal liver metastases and to determine the factors associated with the conversion rate.
First Received: 19 October 2014 Y. Maeda, MD, PhD e-mail: [email protected]
A consecutive series of 210 patients with colorectal liver metastases in a single institute (Hokkaido Cancer Center)
Y. Maeda et al.
treated from 2000–2012 were included in this study. We retrospectively assessed the characteristics of the patients and their survival after the treatments. Primary hepatectomy was performed if technically possible. From 2009, neoadjuvant chemotherapy (6 cycles of FOLFOX) was administrated for primary resectable cases with three or more liver nodules. Resectability was decided based on the size of the remnant liver volume (expected 30 % functional liver remnant after the removal of all metastases), regardless of the number or the size of the liver metastases. Partial resection with a free margin of [1 cm was preferred, and anatomical resection was selected when needed. Even if extrahepatic metastases were present, hepatectomy was performed for the patients whose extrahepatic metastases were also resectable. All detectable lesions were resected to achieve R0 resection, and radiofrequency ablation also was used in some cases for small nodules (\10 mm) located in the deeper liver in some patients. Chemotherapy was administrated to the patients diagnosed with unresectable disease. As first-line chemotherapy, oxaliplatin-based and CPT-11 regimens were mainly used. Before 2008, hepatic artery infusion (HAI) chemotherapy was applied for selected patients. From 2007, an anti-VEGF agent was combined with the regimens of systemic chemotherapy. From 2008, antiEGFR agents were combined with the regimens of systemic chemotherapy for the patients who had wild-type KRAS status. Patients were followed by CT scanning and the resectability was assessed every 2 or 3 months by liver surgeons, medical oncologists, and radiologists. If curative
FIG. 1 Flow diagram of the 210 patients with colorectal liver metastases. Ninety-four patients were initially resectable at the time of diagnosis and underwent hepatectomy (primary resection group). Thirteen of them received neoadjuvant chemotherapy before hepatectomy. The other 116 patients were diagnosed to have unresectable liver metastases, and 104 of them received chemotherapy. Eleven patients converted to be resectable after chemotherapy and underwent hepatectomy (conversion group). Ninety-three patients remained unresectable even after chemotherapy (unresectable group). Twelve patients received best supportive care only
resection became possible, conversion hepatectomy was planed. Chemotherapy was continued until conversion hepatectomy could be performed or until the patient could no longer receive chemotherapy because of tumor progression or due to the adverse effects of the treatments. The overall survival curves from the date of the first treatment for liver metastases (hepatectomy or chemotherapy) were estimated using the Kaplan–Meier method and were analyzed using the log-rank test. To determine the independent relationships between preoperative variables and conversion hepatectomy, Fisher’s exact provability tests were performed. All statistical analyses were performed using the EZR statistical software program.12 P \ 0.05 was considered to be statistical significant. RESULTS A total of 210 patients with colorectal liver metastases were treated between 2000 and 2012. A flow diagram of the 210 patients with colorectal liver metastases is shown in Fig. 1. Ninety-four patients were initially resectable at the time of diagnosis and underwent hepatectomy (primary resection group). Thirteen of them received neoadjuvant chemotherapy before hepatectomy. Sixteen patients also had extrahepatic metastases (lung or others) and underwent resection of both the liver and extrahepatic metastases to achieve R0 resection. The other 116 patients were diagnosed to have unresectable lesions, 104 of whom received chemotherapy. The reasons for unresectability were as follows: liver limited factors in 46 cases (expected functional liver remnant \30 %), the presence of unresectable
210 patients with colorectal liver metastases
94 initially resectable
13 received NAC
116 initially unresectable
104 received chemotherapy
12 BSC only
94 received hepatectomy
11 received conversion hepatectomy
93 remained unresectable
Primaly resection group
Conversion group
Unresectable group
Conversion Hepatectomy for Unresectable Metastases
extrahepatic metastases in 21 cases, and both liver and extrahepatic factors in 37 cases. Eleven patients converted to be resectable after chemotherapy and underwent hepatectomy (conversion group). Ninety-three patients remained unresectable even after chemotherapy (unresectable group). Twelve patients received best supportive care only. The clinical characteristics of the initially resectable (n = 94) and unresectable (n = 116) patients are summarized in Table 1. A significantly higher incidence of extrahepatic metastases, shorter interval to liver metastases, higher number of liver metastatic nodules, larger size of liver metastases, and higher CEA level were observed in initially unresectable group compared to the resectable group. TABLE 1 Patient and tumor characteristics Initially resectable group (n = 94)
Initially unresectable group (n = 116)
P value
Age (years) \75
85
95
C75
9
21
0.11
Gender Female Male Extrahepatic disease
26
51
68
67
Present
16
61
None
78
55
0.022
Factors Predicting Conversion Hepatectomy after Chemotherapy \0.0001
Lymph node metastases N0
27
26
N(?)
63
80
4
10
Colon
59
91
Rectum
35
25
NX
0.42
Location of primary tumor 0.048
Interval to metastases Metachronous
27
5
Synchronous (\12 months)
67
111
81
36
C5 13 Size of largest liver metastasis
80
\0.0001
No. of metastases \5
\5 cm
83
59
C5 cm
11
37
0
20
\20
69
34
C20
25
82
Unknown
\0.0001
\0.0001
CEA level \0.0001
Italic values are used for names of gene or statistical figures (P value) CEA carcinoembryonic antigen
Among the 116 patients in the initially unresectable group, 86 received systemic chemotherapy, 18 received HAI chemotherapy, and 12 received best supportive care only. A weekly high-dose 5FU regimen was used for HAI chemotherapy, and CPT11 also was used as combination treatment in seven of these patients. The first-line protocols for systemic chemotherapy were as follows: CPT-11based regimens in 34 patients, oxaliplatin-based regimens in 40 patients and others (capecitabine, S1, 5FU/LV) in 11 patients. An anti-VEGF agent (bevacizumab) was added in 34 patients, and an anti-EGFR agent (cetuximab or panitumumab) was added in 19 patients with a wildtype KRAS status. The lesions in 11 patients (10.5 %) became resectable after chemotherapy and these subjects subsequently underwent R0 hepatectomy (conversion hepatectomy). The surgical procedures for hepatectomy in the conversion cases were as follows: right lobectomy in five cases (3 patients also received 2–4 wedge resections) and nonanatomical resection in six cases (1–12 wedge resections). One case of surgical mortality within 30 days was seen in primary resection group (1.1 % 1/94); however, no in-hospital death was observed in the conversion patients. The final rate of resection of the liver metastases was 50 % (105/210), including the cases of conversion hepatectomy.
The rate of conversion in the patients with liver-limited metastases was 24 % (11/46). In contrast, no cases of conversion were observed among the patients with extrahepatic metastases (0/58; P \ 0.0001; Table 2). Potential factors predicting conversion hepatectomy in the liver-limited cases are shown in Table 3. The conversion rate was significantly higher in the patients with a negative LN status of the primary lesion (50 %; P = 0.049), the patients treated with hepatic arterial infusion chemotherapy (46 %; P = 0.050), the patients treated with an anti-EGFR agent (44 %; P = 0.013), and the patients exhibiting a complete or partial response (58 %; P \ 0.0001). Long-Term Survival in Patients with Colorectal Liver Metastases Figure 2A shows the results of a Kaplan–Meier analysis of the overall survival in the primary hepatectomy group, the conversion group, the unresectable group, and the best supportive care group. The median follow-up period among the survivors at the last follow-up was 89 months in the conversion group and 56 months in the primary resection group. The overall survival of the conversion group
Y. Maeda et al.
was significantly higher than that of the unresectable group (P \ 0.0001). The predicted 5-year survival rate in the conversion hepatectomy group was 76 %, and five patients survived at least 5 years. Three of these patients survived more than 5 years tumor-free after their final treatments. No patient in the nonconversion group survived more than 5 years.
TABLE 2 Conversion rates according to the presence of extrahepatic metastases Extrahepatic disease
Conversion rate (%)
P value
Present
0
0/58
\0.0001
None (liver limited metastases)
24
11/46
Bold value indicates statistical significance (p \ 0.05)
TABLE 3 Factors predicting conversion hepatectomy in liver limited cases Conversion group (n = 11)
Unresected group (n = 35)
Odds ratio
95 % CI
P value
\75
11
27
–
–
0.17
C75
0
8 2.05
0.43–11.3
0.49
4.77
0.89–28.2
0.049
2.04
0.20–104.4
0.99
–
–
0.11
3.91
0.44–190.4
0.25
–
–
0.99
2.03
0.33–22.4
0.47
0.50
0.098–2.41
0.48
1.42
0.25–7.20
0.71
4.60
0.89–25.9
0.050
0.79
0.14–5.74
0.99
16.3
1.28–935.5
0.013
Age (years)
Gender Female
7
16
Male
4
19
Lymph node metastases N0
6
6
N(?)
5
25
NX
0
4
Histological differentiation of primary tumor Differentiated 10 Undifferentiated
29
1
6
Wild-type
7
8
Mutant
0
5
Unknown
4
22
10
25
1
10
KRAS status
Location of primary tumor Colon Rectum Interval to metastases Metachronous
1
0
10
35
C5
9
24
\5
2
11
5
22
6
13
\20
4
10
C20
7
25
Synchronous (\12 months) No. of metastases
Size of largest liver metastasis C5 cm \5 cm CEA level
Initial treatment for liver metastases HAI
6
7
Systemic chemotherapy
5
28
LOHP or CPT11 based
8
27
Others
3
8
Regimen
Anti-EGFR agenta Used
4
5
Not used
1
23
Conversion Hepatectomy for Unresectable Metastases TABLE 3 continued Conversion group (n = 11)
Unresected group (n = 35)
Odds ratio
95 % CI
P value
–
–
0.29
–
–
£0.0001
Anti-VEGF agenta Used
0
10
Not used
5
18
Response to chemotherapy CR or PR
11
8
SD or PD
0
27
Italic values are used for names of gene or statistical figures (P value). Bolditalic values are statistical significance (P \ 0.05) HAI hepatic artery infusion chemotherapy, LOHP oxialiplatin, CR complete response, PR partial response, SD stable disease, PD progressive disease a
Assessed in patients who underwent systemic chemotherapy only
A
B
Overall survival
1.0
Disease-free survival
0.1 Conversion group (n=11)
0.8
0.8
0.6
0.6 Primary resection group (n=94)
0.4
P
