- Email: [email protected]
Efficacy of repeat hepatectomy for recurrence following curative hepatectomy for colorectal liver metastases: A Retrospective Cohort Study of 128 patients
International Journal of Surgery 36 (2016) 96e103
Contents lists available at ScienceDirect
International Journal of Surgery journal homepage: www.journal-surgery.net
Original Research
Efficacy of repeat hepatectomy for recurrence following curative hepatectomy for colorectal liver metastases: A Retrospective Cohort Study of 128 patients Masakazu Hashimoto, Tsuyoshi Kobayashi*, Kohei Ishiyama, Kentaro Ide, Masahiro Ohira, Hiroyuki Tahara, Shintaro Kuroda, Michinori Hamaoka, Hiroshi Iwako, Masashi Okimoto, Hideki Ohdan Department of Gastroenterological and Transplant Surgery, Applied Life Sciences, Institute of Biomedical & Health Sciences, Hiroshima University, 1-2-3 kasumi, Minami-Ku, Hiroshima, 734-8557, Japan
h i g h l i g h t s Patterns of recurrence after hepatectomy for colorectal liver metastases had important implications for overall survival. Short disease-free interval was the prognostic factors for recurrence of colorectal liver metastases. Short disease-free interval was strongly correlated with high Carcinoembryonic antigen at the hepatectomy.
a r t i c l e i n f o
a b s t r a c t
Article history: Received 15 August 2016 Received in revised form 6 October 2016 Accepted 8 October 2016 Available online 11 October 2016
Background: Despite improvements in surgery and chemotherapy, most patients develop recurrence after initial hepatectomy for colorectal liver metastasis (CRLM). Following initial hepatectomy for CRLM, patterns and surgical management of recurrence have not been widely reported. Materials and Methods: We identified 128 patients who underwent hepatic resection for CRLM between January 2000 and December 2012. Demographics, operative data, site of recurrence, and long-term survival data were collected and analyzed. Patients were stratified into 3 groups based on their site of recurrence as intrahepatic, intra- and extrahepatic, and extrahepatic. In addition, the influence of potential factors on overall survival (OS) in patients with only liver relapse was analyzed through univariate and multivariate analysis. Results: After curative initial hepatectomy, 87 (68.0%) patients had a recurrence: 33 in the intrahepatic group, 11 in the intra- and extrahepatic group, and 43 in the extrahepatic group. The OS for the intra- and extrahepatic group was significantly lower than that for the intrahepatic group. In the intrahepatic group, disease-free interval (DFI) < 12 months and non-repeat hepatectomy were independent poor prognostic factors. Carcinoembryonic antigen (CEA) at the time of hepatectomy was significantly higher in DFI < 12 group than in the DFI 12 group. Conclusion: Patterns of recurrence following initial hepatectomy for CRLM have important implications for OS. In the intrahepatic recurrence group, short DFI was correlated with high CEA at hepatectomy, and was a poor prognostic factor. © 2016 IJS Publishing Group Ltd. Published by Elsevier Ltd. All rights reserved.
Keywords: Colorectal liver metastases Repeat hepatectomy Recurrence Prognostic factor
1. Introduction Colorectal cancer (CRC) is the second leading cause of cancer * Corresponding author. Kasumi, Minami-ku, Hiroshima, 7348551, Japan. E-mail address: [email protected] (T. Kobayashi).
death worldwide [1]. The liver is a common and predominant site for the recurrence of CRC. Synchronous liver metastases are present in about 15e20% of cases at the time of diagnosis of CRC. Metachronous liver metastases occur in about 25e50% of cases within 3 years after the resection of the primary CRC [2e4]. Liver resection has been widely accepted as the only potentially curative treatment
http://dx.doi.org/10.1016/j.ijsu.2016.10.004 1743-9191/© 2016 IJS Publishing Group Ltd. Published by Elsevier Ltd. All rights reserved.
M. Hashimoto et al. / International Journal of Surgery 36 (2016) 96e103
97
for colorectal liver metastases (CRLM) [5]. Although the overall survival outcome of patients with CRLM has improved, recurrence remains common, and occurs in more than 60% of patients within 2 years after the initial hepatectomy [6e9]. Recurrence after the initial hepatectomy was seen at the site of the liver and lung. For the recurrence in the remaining liver after initial hepatectomy, repeat hepatectomy offers long-term survival in selected patients [10,11]. Luo et al. reported that patients should be selected based on the following criteria: less than 1 year of disease-free interval after initial hepatectomy, solitary CRLM, unilobar CRLM, less than 5 cm CRLM (maximum), lack of concomitant extrahepatic metastases, and R0 resection at second hepatectomy [11]. In addition, repeat hepatectomy is not as safe as the initial hepatectomy [12]. For a selected group of patients with metastases limited to the lungs, 5-
Fig. 1. Flow chart demonstrating initial hepatectomy, recurrences, and patterns of recurrence.
Table 1 Patients characteristics of colorectal liver metastases (n ¼ 128). Variables
No recurrence (n ¼ 41)
Location of recurrence
p-value
Intrahepatic (n ¼ 33)
Extrahepatic (n ¼ 43)
Intra- and extrahepatic (n ¼ 11)
Group
1
2
3
4
1 vs 2
2 vs 3, 4
2 vs 4
Age, years Gender Male Female DM Positive Negative Primary tumor characteristics Primary tumor site Colon Rectum T factor ~mp ss~ N factor Positive Negative Lymphatic invasion Positive Negative Venous invasion Positive Negative Histologic type Well Mod. por. etc Unclear Liver metastases characteristics CEA CA19-9 Liver metastases Synchronous Metachronous Tumor number Tumor number Single Multiple Tumor Unilateral Bilateral Maximum tumor size Surgical procedures Minor hepatectomy Major hepatectomy Blood loss Postoperative factors IComplication (Clavian-Dindo) III
66 (42e83)
62 (40e80)
58 (23e77)
67 (52e81)
0.103 0.802
0.433 0.652
0.207 0461
29 (70.7) 12 (29.3)
22 (66.7) 11 (33.3)
24 (55.8) 19 (44.2)
9 (91.8) 2 (18.2) 0.557
0.776
0.391
9 (22.0) 32 (78.1)
5 (15.2) 28 (84.9)
7 (16.3) 36 (83.7)
3 (27.3) 8 (72.7) 0.815
0.381
0.282
26 (63.4) 15 (36.6)
20 (60.6) 13 (39.4)
18 (41.9) 25 (58.1)
9 (81.8) 2 (18.2) 0.725
1.000
1.000
5 (12.2) 36 (87.8)
3 (9.1) 30 (90.9)
3 (7.0) 40 (93.0)
1 (9.1) 10 (90.9) 0.155
0.802
0.408
23 (56.1) 18 (43.9)
24 (72.7) 9 (27.3)
31 (82.1) 12 (27.9)
10 (90.9) 1 (9.1) 0.616
0.614
0.709
27 (65.8) 14 (34.2)
24 (72.7) 9 (27.3)
35 (81.4) 8 (18.6)
7 (63.6) 4 (36.4) 1.000
0.646
1.000
29 (70.7) 12 (29.3)
23 (69.7) 10 (30.3)
27 (62.8) 16 (37.2)
8 (72.7) 3 (27.3) 0.005
0.196
0.194
18 (43.9) 22 (53.7) 1 (2.4)
5 (15.2) 18 (84.8) 0 (0.0)
10 (26.3) 32 (71.4) 1 (2.3)
4 (36.4) 7 (63.6) 0 (0.0)
8.5 (1e221) 13.5 (0e991)
16.9 (0.8e436) 35 (1e884)
22.8 (0.8e1220) 35 (2e2004)
16 (1.3e875.2) 20 (2e5573)
0.012 0.062 0.331
0.905 0.816 1.000
0.860 0.551 0.701
25 (61.0) 16 (39.0) 1 (1e10)
24 (72.7) 9 (27.3) 2 (1e13)
30 (69.7) 13 (30.3) 2 (1e14)
9 (81.8) 2 (18.2) 3 (1e14)
0.020 0.159
0.924 0.503
0.374 0.155
27 (65.9) 14 (34.2)
16 (48.5) 17 (51.5)
19 (44.2) 24 (55.8)
2 (18.2) 9 (81.8) 0.228
0.365
0.380
7 (50.0) 7 (50.0) 25 (8e65)
5 (29.4) 12 (70.6) 28 (12e60)
14 (58.3) 10 (41.7) 32.5 (10e160)
1 (11.1) 8 (88.9) 20 (10e60)
0.604 0.128
0.444 0.142
0.596 0.669
39 (95.1) 2 (4.9) 150 (20e1930)
27 (81.8) 6 (18.2) 380 (30e1000)
27 (62.3) 16 (37.2) 230 (10e2100)
8 (72.3) 3 (27.3) 360 (100e1000)
0.262
0.969
0.862
0.054
0.070
0.170
3 (7.5) 38 (92.5)
8 (25.0) 25 (75.0)
5 (13.2) 38 (86.8)
0 (0.0) 11 (100) 0.238
0.379
0.509
27 (68.9) 10 (24.4) 4 (9.8)
25 (75.8) 5 (15.2) 3 (9.1)
33 (76.7) 6 (14.0) 4 89.3)
9 (81.8) 2 (18.2) 0 (0.0)
Medium (range), n (%).
98
M. Hashimoto et al. / International Journal of Surgery 36 (2016) 96e103
year survival rates of 25%e68% have been reported [13]. Recent data have improved because of increased sensitivity of diagnostic procedures, standardization of follow-up, and better patient selection [14]. In contrast, chemotherapy for patients with CRC has improved recently, and the response rates have significantly increased by oxaliplatin, irinotecan, and target molecular drug. In addition, chemotherapy allows about 10% of patients with unresectable CRLM to be rescued by surgery [15]. However, for some patients, surgery is not performed based on the disease progression and chemotherapy is not continued because of the decreased tolerability after initial hepatectomy. Therefore, the purpose of this study was to investigate recurrence patterns after the initial hepatectomy for CRLM and to determine the prognostic factors after intrahepatic recurrence. Such knowledge might prove useful in selecting the therapy for recurrence of CRLM. 2. Methods Between January 2000 and December 2012, 128 consecutive patients underwent primary hepatic resection with a curative intent for CRLM at the Hiroshima university Hospital, Hiroshima, Japan. We investigated several clinicopathologic variables pertaining to patient characteristics, clinical data, and histopathologic findings, such as sex, age, primary cancer location, lymph node status, timing of first hepatectomy, number of hepatic metastases, tumor diameter, tumor distribution, preoperative serum carcinoembryonic antigen (CEA) level and (CA19-9), extent of liver resection, surgical margin, venous invasion by liver metastases, and bile duct invasion. The demographic data were collected from a prospectively maintained database. All tumors were identified routinely by contrast-enhanced computed tomography (CT) during hepatic arteriography and portography, magnetic resonance imaging (MRI), and ultrasonography (US), which were performed as preoperative studies in patients considered for the hepatectomy. The patients with resectable CRLM were treated by immediate resection or neoadjuvant chemotherapy based on the surgeon's discretion, whereas those with unresectable CRLM were treated with chemotherapy including oxaliplatin or irinotecan. Hepatectomy was planned after reassessment. It was indicated in cases where all tumors could be removed with clear margins with the remnant liver function preserved, and hepatic resection could be performed safely [16,17]. All CRLM cases were assessed for resectability by a liver surgeon, using the same criterion. During surgery, abdominal exploration and intraoperative ultrasonography were routinely used to confirm resectability, with non-anatomical liver resection with adequate free margins by the standard procedure. Major hepatectomy was defined as a resection of three or more segments as described by Couinaud, and minor hepatectomy was defined as wedge, segmental, and sectional resections. Postoperative complications were classified according to the Clavien-Dindo classification, and R0 resection was defined as no microscopic evidence of tumor at the resection margin. After initial hepatectomy, patients were routinely followed every 3 months. The levels of serum CEA and CA19-9 were measured at every outpatient clinic visit. Postoperative adjuvant chemotherapy in patients with R0 resection was not administered routinely, except in those recommended by the surgeon. Surveillance imaging, including ultrasonography and CT of the abdomen and thorax, was conducted every 3 months during the first 2 years and every 6 months until the disease-free interval exceeded 5 years.
The first relapse was defined as the site diagnosed by imaging after the initial hepatic resection, and we classified it as intrahepatic group, intra- and extrahepatic group, and extrahepatic group. The recurrent CRLM were treated with the same indication as the primary CRLM. Chemotherapy response of liver metastases was measured according to the revised response evaluation criteria in solid tumors (RECIST) criteria [18]. Categorical variables were compared using the Chi-square test, while continuous variables were compared using the ManneWhitney U-test. Overall and progression-free survival after hepatic resection were calculated by the KaplaneMeier method and compared by the log-rank test. Overall survival (OS) rates and disease-free survival (DFS) were calculated using the KaplaneMeier method. Differences between curves were assessed according to the log-rank test. Independent prognostic factors were assessed using the multivariate analysis. Cox proportional hazards model among the variables was found to be significant on univariate analysis. Differences of P < 0.05 were considered significant. All statistical analyses were performed using the JMP 10 for Windows (SAS Institute Japan). 3. Results 3.1. Comparison of clinicopathological characteristics and prognosis according to the recurrence patterns after initial hepatectomy The median follow-up for survivors as measured from the time of hepatectomy was 47.8 months. At the last follow-up, 52 patients (40.6%) had died of disease, 4 (3.1%) had died of other causes, 31 (24.2%) were alive with disease, and 41 (32.0%) were alive with no evidence of disease. Overall, 87 patients (68.0%) experienced a recurrence of disease at the last follow-up. Among the patients with recurrence, the median follow-up from the time of the initial recurrence was 27.3 months. In primary hepatectomy, 88 patients had synchronous
Fig. 2. Kaplan-Meier curves for patients without recurrence, and for those with intrahepatic, intra- and extrahepatic, and extrahepatic recurrence after initial hepatectomy, showing overall survival (A) and disease-free survival (B). Overall survival rates at 5 years were 100.0%, 46.6%, 0.0%, and 39.8%, respectively. Disease-free survival rates at 2 years were 100.0%, 6.1%, 0.0%, and 7.1%, respectively.
M. Hashimoto et al. / International Journal of Surgery 36 (2016) 96e103
metastases of CRLM and 40 had metachronous metastases of CRLM. The average number of tumors was 2.6 and average maximum size of tumors was 34.4 mm. Of the patients, 94 (73.4%) were administered adjuvant chemotherapy after primary hepatectomy and had no recurrence. After hepatectomy, 33 patients had only intrahepatic metastases, 11 had intra- and extrahepatic metastases, and 43 had only extrahepatic metastases (Fig. 1). On comparing patients in the no recurrence group and those in the intrahepatic group, those in the no recurrence group showed more well-differentiated histologic type of primary tumor than the patients from the intrahepatic group. In addition, the CEA levels and tumor numbers at the primary hepatectomy were significantly lower in the no recurrence group than in the intrahepatic group (Table 1). In contrast, no significant difference was found between the recurrence site of the intrahepatic group and the extrahepatic group as well as between the intrahepatic group and the intra- and extrahepatic group. The OS rates of the intra- and extrahepatic group were significantly lower than that of the intrahepatic group (3- and 5-year
Table 2 Overall survival rates according to clinicopathological factors. Overall survival (%)
All cases (n ¼ 33) Age, years < 65 (n ¼ 18) 65 (n ¼ 15) Gender Male (n ¼ 22) Female (n ¼ 11) Primary tumor Colon (n ¼ 20) Rectum (n ¼ 13) Histological grading Well (n ¼ 5) Moderately or poorly (n ¼ 28) Lymphatic invasion Yes (n ¼ 23) No (n ¼ 9) Venous invasion Yes (n ¼ 22) No (n ¼ 10) Depth of tumor invasion sm, mp (n ¼ 3) ss~ (n ¼ 30) Lymph node metastases Positive (n ¼ 24) Negative (n ¼ 9) Tumor size (max; mm) < 30 (n ¼ 17) 30 (n ¼ 16) Number of tumors Single (n ¼ 16) Multiple (n ¼ 17) CEA < 20 (n ¼ 19) 20 (n ¼ 14) CA19-9 < 40 (n ¼ 18) 40 (n ¼ 15) Type of hepatectomy Seggmentectomy or less (n ¼ 27) Sectionectomy or more (n ¼ 6) Disease free interval (month) < 12 (n ¼ 22) 12 (n ¼ 11) Repeat hepatectomy Yes (n ¼ 17) No (n ¼ 16)
3-year
5-year
p value
63.1
46.6
e 0.311
70.8 56.0
52.5 42.0 0.984
60.9 68.2
49.8 40.9
61.5 67.7
46.2 47.4
30.0 68.9
30.0 50.1
64.1 59.3
41.2 59.3
61.1 67.5
40.7 56.3
66.7 63.3
66.7 44.2
59.2 74.1
36.9 74.1
65.3 59.1
48.9 44.3
65.7 61.4
57.5 34.5
72.9 43.6
60.7 21.8
69.6 54.4
61.9 29.2
67.7 41.7
48.8 41.7
0.761
99
survival rates of 20.0% and 0.0%, versus 63.1% and 46.6%, respectively) (Fig. 2). In addition, the patients from the “no recurrence” group did not die. 3.2. Analysis in the intrahepatic recurrence after primary hepatectomy Table 2 summarizes the results of the univariate analyses for OS rates according to clinicopathological factors in intrahepatic recurrence after hepatectomy. Disease free interval <12 months and non-repeat hepatectomy were significant adverse prognostic factors for OS in patients with the intrahepatic recurrence after primary hepatectomy (Fig. 3). Table 3 summarizes the results of the comparison of clinicopathological characteristics between the recurrence <12 months (n ¼ 22) and the recurrence 12 months (n ¼ 11). CEA levels at the primary hepatectomy were significantly higher in the recurrence <12 months group than in the recurrence 12 months group. The percentages of patients with venous invasion of the primary tumor and CA19-9 levels at the primary hepatectomy tended to be higher in the recurrence <12 months group than in the recurrence 12 months group. Table 4 summarizes the results of the comparison of clinicopathological characteristics between patients with repeat hepatectomy (n ¼ 19) and those without repeat hepatectomy (n ¼ 14). The disease-free interval (DFI) after the primary hepatectomy was significantly shorter in patients with repeat hepatectomy than that in patients without repeat hepatectomy. 3.3. Analyses of treatment of patients with DFI <12 months after initial hepatectomy Twenty-two patients with DFI <12 months after primary
0.159
0.655
0.410
0.942
0.343
0.881
0.330
0.126
0.285
0.354
0.001 40.9 100
18.2 90.0
87.4 31.3
80.7 0.0
<0.001
Fig. 3. (A) Kaplan-Meier curves for patients with DFI < 12 months and those with DFI 12 months after initial hepatectomy showing overall survival. OS rates with DFI <12 months at 3 and 5 years (40.9% and 18.2%, respectively) were significantly lower than for those with DFI 12 months (100.0% and 90.0%, respectively) (p ¼ 0.001). (B) Kaplan-Meier curves for patients with and without repeat hepatectomy after initial hepatectomy showing overall survival. OS rates in the non-repeat hepatectomy group at 3 and 5 years (31.3% and 0.0%, respectively) were significantly lower than those for the repeat hepatectomy group (87.4% and 80.7%, respectively) (p < 0.001).
100
M. Hashimoto et al. / International Journal of Surgery 36 (2016) 96e103
Table 3 Comparison of patients characteristics with intrahepatic recurrences accoding to DFI. Variables Patients characteristics Age, years Gender Male Female DM Positive Negative Primary tumor characteristics Primary tumor site Colon Rectum T factor ~mp ss~ N factor Positive Negative Lymphatic invasion Positive Negative Venous invasion Positive Negative Histologic type well mod. por. etc Liver metastases characteristics CEA CA19-9 Preoperative chemotherapy Tumor number Tumor number Single Multiple Tumor hemi bi Maximum tumor size Surgical procedures Minor hepatectomy Major hepatectomy Blood loss Postoperative factors Complication (clavian-Dindo) III
<12 months (n ¼ 22)
12 months (n ¼ 11)
p-value
65 (40e80)
60 (50e75)
0.504 0.709
14 (63.6) 8 (36.4)
8 (72.7) 3 (27.3)
3 (13.6) 19 (86.4)
2 (18.2) 9 (81.8)
13 (59.1) 9 (40.9)
7 (63.6) 4 (36.4)
2 (9.1) 20 (90.9)
1 (9.1) 10 (90.9)
17 (77.3) 5 (22.7)
7 (63.6) 4 (36.4)
16 (76.2) 5 (23.8)
7 (63.6) 4 (36.4)
17 (81.0) 4 (19.1)
5 (45.5) 6 (54.6)
5 (22.7) 17 (77.3)
0 (0.0) 11 (100)
49.0 (0.8e436) 65.5 (1e884)
7.5 (3.1e21.7) 12 (3e53)
0.011 0.059
3 (1e13)
1 (1e5)
0.127 0.282
9 (40.9) 13 (59.1)
7 (63.6) 4 (36.4)
3 (23.1) 10 (76.9) 22 (14e60)
2 (50.0) 2 (50.0) 38 (12e50)
17 (77.3) 5 (22.7) 380 (30e1000)
10 (90.9) 1 (9.1) 190 (70e520)
5 (23.8) 16 (76.2)
3 (27.3) 8 (72.7)
27 (68.9) 10 (24.4) 4 (9.8)
25 (75.8) 5 (15.2) 3 (9.1)
7 (31.8) 15 (63.2)
10 (90.9) 1 (9.1)
1.000
1.000
1.000
0.438
0.681
0.056
0.538
0.193 0.637 0.341 1.000
0.002
Medium (range), n (%).
hepatectomy were divided into repeat hepatectomy group (n ¼ 8) and non-repeat hepatectomy group (n ¼ 14) (Fig. 4). Among the repeat hepatectomy group, 4 patients had chemotherapy before hepatectomy, whereas, 4 underwent immediate repeat hepatectomy. Table 5 shows the comparison between patients who underwent hepatectomy and those who did not among the patients with DFI <12 months. At the time of the intrahepatic recurrence, the number of intrahepatic recurrences was significantly lower in the repeat hepatectomy group than that in the non-repeat hepatectomy group. In the repeat hepatectomy group, 3 (75%) patients responded to chemotherapy with Stable Disease (SD) or better, but 1 (25%) patient had no response. However, in the non-repeat hepatectomy group, 1
(7.1%) patient had a response, but 13 (92.9%) had no response. The response to chemotherapy for intrahepatic recurrence was significantly lower in the non-repeat hepatectomy group than in the repeat hepatectomy group. 4. Discussion We investigated the patterns of recurrences that occurred after the initial hepatic resection of colorectal hepatic metastases because the recurrence after initial hepatic resection was about 70%. In addition, we particularly studied the recurrence that occurred in the liver for improving the treatment of intrahepatic recurrence after the initial hepatectomy.
M. Hashimoto et al. / International Journal of Surgery 36 (2016) 96e103
101
Table 4 Comparison of characteristics of patients with intrahepatic recurrences accoding to with or without repeat hepatectomy. Variables Patients characteristics Age, years Gender Male Female DM Positive Negative Primary tumor characteristics Primary tumor site Colon Rectum T factor ~mp ss~ N factor Positive Negative Lymphatic invasion Positive Negative Venous invasion Positive Negative Histologic type Well Mod. por. etc Liver metastases characteristics CEA CA19-9 Preoperative chemotherapy Tumor number Tumor number Single Multiple Tumor Hemi Bi Maximum tumor size Surgical procedures Minor hepatectomy Major hepatectomy Blood loss Postoperative factors Complication (clavian-Dindo) III
Repeat hepatectomy (n ¼ 17)
Non-repeat hepatectomy (n ¼ 16)
p-value
60 (40e80)
66 (45e79)
0.113 0.721
12 (70.6) 5 (29.4)
10 (62.5) 6 (37.5)
3 (17.7) 14 (82.4)
2 (12.5) 14 (87.5)
10 (58.8) 7 (41.2)
10 (62.5) 6 (37.5)
3 (17.7) 14 (82.4)
0 (0.0) 16 (100)
12 (70.6) 5 (29.4)
12 (75.0) 4 (25.0)
12 (70.6) 5 (29.4)
11 (73.3) 4 (26.7)
10 (58.8) 7 (41.2)
12 (80.0) 4 (20.0)
1 (5.9) 16 (94.1)
4 (25.0) 12 (75.0)
11.2 (2.7e235) 24 (1e884)
49.0 (0.8e436) 44 (2e505)
0.249 0.438
1 (1e13)
3 (1e13)
0.127 0.732
9 (52.9) 8 (47.1)
7 (43.8) 9 (56.3)
4 (50.0) 4 (50.0) 28 (12e60)
1 (11.1) 8 (88.9) 30 (14e50)
15 (88.2) 2 (11.8) 400 (30e1000)
12 (75.0) 4 (25.0) 380 (40e650)
3 (17.7) 14 (82.4)
4 (26.7) 11 (73.3)
14 (82.4) 3 (17.7) 0 (0.0) 14.3 (2.8e51.6)
11 (68.8) 2 (12.5) 3 (18.8) 6.3 (1.3e15.7)
1.000
1.000
0.227
1.000
1.000
0.265
0.175
0.131
0.339 0.398 0.341 0.678
1.000
0.006
Medium (range), n (%). DFI; Disease Free Interval.
Several studies reported the pattern of recurrence after initial hepatectomy in patients with CRLM [7,19]. These results were similar to those of our study. Our finding of recurrence within 2 years was also similar to that of other studie [7]. In the current study, we showed that DFI and non-repeat hepatectomy were risk factors in OS for the prognosis improvement in the intrahepatic metastases. A DFI greater than 12 months between initial hepatectomy and recurrence for CRLM has been associated with improved long-term survival in a number of studies [20e22]. In the current study, when the characteristics of the patients of DFI <12months and DFI >12months were compared, the patients of DFI <12 months had a higher serum level of CEA and underwent a nonrepeat hepatectomy. Similarly, when the characteristics of the patients in the repeat hepatectomy and without repeat hepatectomy
group were compared, those in the repeat hepatectomy group had longer DFI. A DFI <12 months was believed to be the reason for the patients' not undergoing repeat hepatectomy. Based on these results, examining the treatment choice for a better prognosis and developing a strategy to prolong the DFI is important. For hepatectomy of CRLM, the criteria for resectability of colorectal liver metastases have been significantly revised and developed recently. Timothy M. reported that the feasibility of hepatic resection should be not only be determined by the old criteria, such as the number of metastases, the size of metastases, and the extrahepatic disease, but also by R0 resection and adequate remnant liver function [23]. Moreover, repeat hepatectomy is feasible in selected patients and seems justified, as it may result in prolonged survival with acceptable rates of morbidity and
102
M. Hashimoto et al. / International Journal of Surgery 36 (2016) 96e103
Fig. 4. Kaplan-Meier curves for patients with and without repeat hepatectomy in the group with DFI <12 months after initial hepatectomy, showing overall survival. OS rates in the non-repeat hepatectomy group at 3 and 5 years (15.3% and 0.0%, respectively) were significantly lower than those for the repeat hepatectomy group (72.9% and 38.9%, respectively) (p ¼ 0.004).
mortality [22,23]. In addition, repeat hepatectomy such as 2nd, 3rd, and 4th hepatectomy results were similar to those seen after the initial hepatectomy [24]. For the recurrence after hepatectomy, the strategy of neoadjuvant chemotherapy in patients with resectable CRLM is considered because it has advantages such as an improved progression-free survival, the evaluation of chemo-responsiveness, and selection of surgery. However, for delayed surgery, cases have been reported where unresectable tumors and tumor progression occurred while treating with neo-adjuvant chemotherapy. In the current study, the same protocol could be continued in most patients; therefore, decreasing the dose of chemotherapy was necessary. As a result, only 22.2% patients had achieved the effect of the chemotherapy of Stable Disease (SD) or more. Rolf Sauer et al. proved that the toxic effect by the chemotherapy occurred more in the postoperative treatment group than in the preoperative treatment group for patients with locally advanced rectal cancer [25] and the success rates were higher in the preoperative treatment group than in the postoperative treatment group. Similarly, for the
Table 5 Patients characteristics of only colorectal liver metastases and recurrence within 12 months after hepatectomy (n ¼ 22). Variables
Repeat hepatectomy (n ¼ 8)
Non-repeat hepatectomy (n ¼ 14)
p-value
68.7 (2.7e436) 89 (1e884) 3 (1e13)
49.0 (0.8e396) 44 (2e505) 3 (1e13)
0.246 0.562 0.803 1.000
3 (37.5) 5 (62.5) 22 (15e60)
6 (42.9) 8 (57.1) 25 (14e50)
7.2 (1e87.5) 11 (1e40) 1 (1e4)
14.2 (9.6e116) 13 (2e92) 3.5 (1e7)
5 (62.5) 3 (37.5) 0 (0.0) 14.5 (10e37)
1 (7.1) 9 (64.3) 4 (28.6) 18 (10e35)
1 2 1 4
1 (7.1) 0 (0.0) 13 (92.9) 0 (0.0)
At the initial hepatectomy CEA CA19-9 Number of tumor Number of tumor Single Multiple Maximum tumor size At the intrahepatic recurrence CEA CA19-9 Number of tumor Number of tumor Single Multiple Unclear Maximum tumor size Chemotherapy after recurrence PR SD PD Non Medium (range), n (%).
(12.5) (25.0) (12.5) (50.0)
0.391 0.194 0.685 0.029 0.043
0.348 0.019
patients with CRLM, the tolerability for chemotherapy is considered to be higher in preoperative treatment group than in postoperative treatment group. In other words, controlling recurrences is thought to be difficult because of the low tolerability and inadequate amount of drug after hepatectomy. Therefore, we suggest that repeat hepatectomy is aggressively performed if the relapse tumor after initial hepatectomy is resectable. This study has several limitations to consider. First, it is a singleinstitution study and retrospective in nature. Second, the study is limited by the sample size and the long study period. Additionally, it is likely that other factors contributed to the decision for repeat hepatectomy or palliative therapy. Despite these limitations, the current results help to further clarify the characteristics of patients with recurrence following initial hepatectomy for CRLM. 5. Conclusion Patterns of recurrence after initial hepatectomy for CRLM have important implications for OS. In the intrahepatic recurrence group, short DFI was a poor prognostic factor; moreover, short DFI was correlated with CEA and CA19-9 at hepatectomy and was associated with venous invasion by the primary tumor. Ethical approval This study was reviewed and approved by the institutional review board of Hiroshima university Hospital. Sources of funding There is no source of funding for our research. Authors' contributions M. Hashimoto: study design, data collection, data analysis, drafting and revising manuscript. T. Kobayashi: study design, drafting and revising manuscript. K.Ishiyama: management of patients. K. Ide: management of patients. M. Ohira: management of patients. H. Tahara: management of patients. S. Kuroda: management of patients, data analysis; M. Hamaoka: data collection. H. Iwako: data collection. M. Okimoto: data collection. H. Ohdan: study design, drafting and revising manuscript. Conflict of interest The authors declare that they have no conflicts of interest concerning this article. Guarantor Hideki Ohdan [email protected] Department of Gastroenterological and Transplant Surgery, Applied Life Sciences, Institute of Biomedical & Health Sciences, Hiroshima University, 1-2-3 kasumi, Minami-Ku, Hiroshima 7348557, Japan. Phone numbers: þ81-082-257-5222. Acknowledgment The authors thank Dr. M Hattori (Hiroshima University, Japan) for his assistance with the statistical analysis.
M. Hashimoto et al. / International Journal of Surgery 36 (2016) 96e103
References [1] R. Siegel, D. Naishadham, A. Jemal, Cancer statistics, 2013, CA Cancer J. Clin. 63 (2013) 11e30. [2] R. Lochan, S.A. White, D.M. Manas, Liver resection for colorectal liver metastasis, Surg. Oncol. 16 (2007) 33e45. [3] J. Leporrier, J. Maurel, L. Chiche, et al., A population-based study of the incidence, management and prognosis of hepatic metastases from colorectal cancer, Br. J. Surg. 93 (2006) 465e474. [4] A. Altendorf-Hofmann, J. Scheele, A critical review of the major indicators of prognosis after resection of hepatic metastases from colorectal carcinoma, Surg. Oncol. Clin. N. Am. 12 (2003) 165e192. [5] P.C. Simmonds, J.N. Primrose, J.L. Colquitt, et al., Surgical resection of hepatic metastases from colorectal cancer: a systematic review of published studies, Br. J. Cancer 94 (2006) 982e999. [6] H. Yamada, S. Kondo, S. Okushiba, et al., Analysis of predictive factors for recurrence after hepatectomy for colorectal liver metastases, World J. Surg. 25 (2001) 1129e1133. [7] M.C. de Jong, C. Pulitano, D. Ribero, et al., Rates and patterns of recurrence following curative intent surgery for colorectal liver metastasis: an international multi-institutional analysis of 1669 patients, Ann. Surg. 250 (2009) 440e448. [8] E.K. Abdalla, J.N. Vauthey, L.M. Ellis, et al., Recurrence and outcomes following hepatic resection, radiofrequency ablation, and combined resection/ablation for colorectal liver metastases, Ann. Surg. 239 (2004) 818e825 discussion 825e817. [9] M. D'Angelica, P. Kornprat, M. Gonen, et al., Effect on outcome of recurrence patterns after hepatectomy for colorectal metastases, Ann. Surg. Oncol. 18 (2011) 1096e1103. [10] H.H. Eker, Repeat hepatectomy for recurrent colorectal metastases, Br. J. Surg. 100 (2013) 808e818. The British journal of surgery 2013; 100: 818-819. [11] L.X. Luo, Z.Y. Yu, J.W. Huang, H. Wu, Selecting patients for a second hepatectomy for colorectal metastases: an systemic review and meta-analysis, Eur. J. Surg. Oncol. J. Eur. Soc. Surg. Oncol. Br. Assoc. Surg. Oncol. 40 (2014) 1036e1048. [12] Y. Mise, K. Hasegawa, J. Shindoh, et al., The feasibility of third or more repeat
[13]
[14]
[15]
[16] [17]
[18]
[19]
[20] [21] [22]
[23] [24]
[25]
103
hepatectomy for recurrent hepatocellular carcinoma, Ann. Surg. 262 (2015) 347e357. M. Gonzalez, A. Poncet, C. Combescure, et al., Risk factors for survival after lung metastasectomy in colorectal cancer patients: a systematic review and meta-analysis, Ann. Surg. Oncol. 20 (2013) 572e579. M.G. Zampino, P. Maisonneuve, P.S. Ravenda, et al., Lung metastases from colorectal cancer: analysis of prognostic factors in a single institution study, Ann. Thorac. Surg. 98 (2014) 1238e1245. R. Adam, V. Delvart, G. Pascal, et al., Rescue surgery for unresectable colorectal liver metastases downstaged by chemotherapy: a model to predict long-term survival, Ann. Surg. 240 (2004) 644e657 discussion 657e648. M. Makuuchi, T. Kosuge, T. Takayama, et al., Surgery for small liver cancers, Seminars Surg. Oncol. 9 (1993) 298e304. T. Itamoto, K. Katayama, H. Nakahara, et al., Autologous blood storage before hepatectomy for hepatocellular carcinoma with underlying liver disease, Br. J. Surg. 90 (2003) 23e28. E.A. Eisenhauer, P. Therasse, J. Bogaerts, et al., New response evaluation criteria in solid tumours: revised RECIST guideline (version 1.1), Eur. J. cancer 45 (2009) 228e247. Y. Mise, H. Imamura, T. Hashimoto, et al., Cohort study of the survival benefit of resection for recurrent hepatic and/or pulmonary metastases after primary hepatectomy for colorectal metastases, Ann. Surg. 251 (2010) 902e909. R. Adam, G. Pascal, D. Azoulay, et al., Liver resection for colorectal metastases: the third hepatectomy, Ann. Surg. 238 (2003) 871e883 discussion 883e874. G. Sugawara, M. Isogai, Y. Kaneoka, et al., Repeat hepatectomy for recurrent colorectal metastases, Surg. Today 35 (2005) 282e289. A. Sa Cunha, C. Laurent, A. Rault, et al., A second liver resection due to recurrent colorectal liver metastases, Archives Surg. 142 (2007) 1144e1149 discussion 1150. T.M. Pawlik, R.D. Schulick, M.A. Choti, Expanding criteria for resectability of colorectal liver metastases, Oncology 13 (2008) 51e64. A. Antoniou, R.E. Lovegrove, H.S. Tilney, et al., Meta-analysis of clinical outcome after first and second liver resection for colorectal metastases, Surgery 141 (2007) 9e18. R. Sauer, H. Becker, W. Hohenberger, et al., Preoperative versus postoperative chemoradiotherapy for rectal cancer, N. Engl. J. Med. 351 (2004) 1731e1740.
Contents lists available at ScienceDirect
International Journal of Surgery journal homepage: www.journal-surgery.net
Original Research
Efficacy of repeat hepatectomy for recurrence following curative hepatectomy for colorectal liver metastases: A Retrospective Cohort Study of 128 patients Masakazu Hashimoto, Tsuyoshi Kobayashi*, Kohei Ishiyama, Kentaro Ide, Masahiro Ohira, Hiroyuki Tahara, Shintaro Kuroda, Michinori Hamaoka, Hiroshi Iwako, Masashi Okimoto, Hideki Ohdan Department of Gastroenterological and Transplant Surgery, Applied Life Sciences, Institute of Biomedical & Health Sciences, Hiroshima University, 1-2-3 kasumi, Minami-Ku, Hiroshima, 734-8557, Japan
h i g h l i g h t s Patterns of recurrence after hepatectomy for colorectal liver metastases had important implications for overall survival. Short disease-free interval was the prognostic factors for recurrence of colorectal liver metastases. Short disease-free interval was strongly correlated with high Carcinoembryonic antigen at the hepatectomy.
a r t i c l e i n f o
a b s t r a c t
Article history: Received 15 August 2016 Received in revised form 6 October 2016 Accepted 8 October 2016 Available online 11 October 2016
Background: Despite improvements in surgery and chemotherapy, most patients develop recurrence after initial hepatectomy for colorectal liver metastasis (CRLM). Following initial hepatectomy for CRLM, patterns and surgical management of recurrence have not been widely reported. Materials and Methods: We identified 128 patients who underwent hepatic resection for CRLM between January 2000 and December 2012. Demographics, operative data, site of recurrence, and long-term survival data were collected and analyzed. Patients were stratified into 3 groups based on their site of recurrence as intrahepatic, intra- and extrahepatic, and extrahepatic. In addition, the influence of potential factors on overall survival (OS) in patients with only liver relapse was analyzed through univariate and multivariate analysis. Results: After curative initial hepatectomy, 87 (68.0%) patients had a recurrence: 33 in the intrahepatic group, 11 in the intra- and extrahepatic group, and 43 in the extrahepatic group. The OS for the intra- and extrahepatic group was significantly lower than that for the intrahepatic group. In the intrahepatic group, disease-free interval (DFI) < 12 months and non-repeat hepatectomy were independent poor prognostic factors. Carcinoembryonic antigen (CEA) at the time of hepatectomy was significantly higher in DFI < 12 group than in the DFI 12 group. Conclusion: Patterns of recurrence following initial hepatectomy for CRLM have important implications for OS. In the intrahepatic recurrence group, short DFI was correlated with high CEA at hepatectomy, and was a poor prognostic factor. © 2016 IJS Publishing Group Ltd. Published by Elsevier Ltd. All rights reserved.
Keywords: Colorectal liver metastases Repeat hepatectomy Recurrence Prognostic factor
1. Introduction Colorectal cancer (CRC) is the second leading cause of cancer * Corresponding author. Kasumi, Minami-ku, Hiroshima, 7348551, Japan. E-mail address: [email protected] (T. Kobayashi).
death worldwide [1]. The liver is a common and predominant site for the recurrence of CRC. Synchronous liver metastases are present in about 15e20% of cases at the time of diagnosis of CRC. Metachronous liver metastases occur in about 25e50% of cases within 3 years after the resection of the primary CRC [2e4]. Liver resection has been widely accepted as the only potentially curative treatment
http://dx.doi.org/10.1016/j.ijsu.2016.10.004 1743-9191/© 2016 IJS Publishing Group Ltd. Published by Elsevier Ltd. All rights reserved.
M. Hashimoto et al. / International Journal of Surgery 36 (2016) 96e103
97
for colorectal liver metastases (CRLM) [5]. Although the overall survival outcome of patients with CRLM has improved, recurrence remains common, and occurs in more than 60% of patients within 2 years after the initial hepatectomy [6e9]. Recurrence after the initial hepatectomy was seen at the site of the liver and lung. For the recurrence in the remaining liver after initial hepatectomy, repeat hepatectomy offers long-term survival in selected patients [10,11]. Luo et al. reported that patients should be selected based on the following criteria: less than 1 year of disease-free interval after initial hepatectomy, solitary CRLM, unilobar CRLM, less than 5 cm CRLM (maximum), lack of concomitant extrahepatic metastases, and R0 resection at second hepatectomy [11]. In addition, repeat hepatectomy is not as safe as the initial hepatectomy [12]. For a selected group of patients with metastases limited to the lungs, 5-
Fig. 1. Flow chart demonstrating initial hepatectomy, recurrences, and patterns of recurrence.
Table 1 Patients characteristics of colorectal liver metastases (n ¼ 128). Variables
No recurrence (n ¼ 41)
Location of recurrence
p-value
Intrahepatic (n ¼ 33)
Extrahepatic (n ¼ 43)
Intra- and extrahepatic (n ¼ 11)
Group
1
2
3
4
1 vs 2
2 vs 3, 4
2 vs 4
Age, years Gender Male Female DM Positive Negative Primary tumor characteristics Primary tumor site Colon Rectum T factor ~mp ss~ N factor Positive Negative Lymphatic invasion Positive Negative Venous invasion Positive Negative Histologic type Well Mod. por. etc Unclear Liver metastases characteristics CEA CA19-9 Liver metastases Synchronous Metachronous Tumor number Tumor number Single Multiple Tumor Unilateral Bilateral Maximum tumor size Surgical procedures Minor hepatectomy Major hepatectomy Blood loss Postoperative factors IComplication (Clavian-Dindo) III
66 (42e83)
62 (40e80)
58 (23e77)
67 (52e81)
0.103 0.802
0.433 0.652
0.207 0461
29 (70.7) 12 (29.3)
22 (66.7) 11 (33.3)
24 (55.8) 19 (44.2)
9 (91.8) 2 (18.2) 0.557
0.776
0.391
9 (22.0) 32 (78.1)
5 (15.2) 28 (84.9)
7 (16.3) 36 (83.7)
3 (27.3) 8 (72.7) 0.815
0.381
0.282
26 (63.4) 15 (36.6)
20 (60.6) 13 (39.4)
18 (41.9) 25 (58.1)
9 (81.8) 2 (18.2) 0.725
1.000
1.000
5 (12.2) 36 (87.8)
3 (9.1) 30 (90.9)
3 (7.0) 40 (93.0)
1 (9.1) 10 (90.9) 0.155
0.802
0.408
23 (56.1) 18 (43.9)
24 (72.7) 9 (27.3)
31 (82.1) 12 (27.9)
10 (90.9) 1 (9.1) 0.616
0.614
0.709
27 (65.8) 14 (34.2)
24 (72.7) 9 (27.3)
35 (81.4) 8 (18.6)
7 (63.6) 4 (36.4) 1.000
0.646
1.000
29 (70.7) 12 (29.3)
23 (69.7) 10 (30.3)
27 (62.8) 16 (37.2)
8 (72.7) 3 (27.3) 0.005
0.196
0.194
18 (43.9) 22 (53.7) 1 (2.4)
5 (15.2) 18 (84.8) 0 (0.0)
10 (26.3) 32 (71.4) 1 (2.3)
4 (36.4) 7 (63.6) 0 (0.0)
8.5 (1e221) 13.5 (0e991)
16.9 (0.8e436) 35 (1e884)
22.8 (0.8e1220) 35 (2e2004)
16 (1.3e875.2) 20 (2e5573)
0.012 0.062 0.331
0.905 0.816 1.000
0.860 0.551 0.701
25 (61.0) 16 (39.0) 1 (1e10)
24 (72.7) 9 (27.3) 2 (1e13)
30 (69.7) 13 (30.3) 2 (1e14)
9 (81.8) 2 (18.2) 3 (1e14)
0.020 0.159
0.924 0.503
0.374 0.155
27 (65.9) 14 (34.2)
16 (48.5) 17 (51.5)
19 (44.2) 24 (55.8)
2 (18.2) 9 (81.8) 0.228
0.365
0.380
7 (50.0) 7 (50.0) 25 (8e65)
5 (29.4) 12 (70.6) 28 (12e60)
14 (58.3) 10 (41.7) 32.5 (10e160)
1 (11.1) 8 (88.9) 20 (10e60)
0.604 0.128
0.444 0.142
0.596 0.669
39 (95.1) 2 (4.9) 150 (20e1930)
27 (81.8) 6 (18.2) 380 (30e1000)
27 (62.3) 16 (37.2) 230 (10e2100)
8 (72.3) 3 (27.3) 360 (100e1000)
0.262
0.969
0.862
0.054
0.070
0.170
3 (7.5) 38 (92.5)
8 (25.0) 25 (75.0)
5 (13.2) 38 (86.8)
0 (0.0) 11 (100) 0.238
0.379
0.509
27 (68.9) 10 (24.4) 4 (9.8)
25 (75.8) 5 (15.2) 3 (9.1)
33 (76.7) 6 (14.0) 4 89.3)
9 (81.8) 2 (18.2) 0 (0.0)
Medium (range), n (%).
98
M. Hashimoto et al. / International Journal of Surgery 36 (2016) 96e103
year survival rates of 25%e68% have been reported [13]. Recent data have improved because of increased sensitivity of diagnostic procedures, standardization of follow-up, and better patient selection [14]. In contrast, chemotherapy for patients with CRC has improved recently, and the response rates have significantly increased by oxaliplatin, irinotecan, and target molecular drug. In addition, chemotherapy allows about 10% of patients with unresectable CRLM to be rescued by surgery [15]. However, for some patients, surgery is not performed based on the disease progression and chemotherapy is not continued because of the decreased tolerability after initial hepatectomy. Therefore, the purpose of this study was to investigate recurrence patterns after the initial hepatectomy for CRLM and to determine the prognostic factors after intrahepatic recurrence. Such knowledge might prove useful in selecting the therapy for recurrence of CRLM. 2. Methods Between January 2000 and December 2012, 128 consecutive patients underwent primary hepatic resection with a curative intent for CRLM at the Hiroshima university Hospital, Hiroshima, Japan. We investigated several clinicopathologic variables pertaining to patient characteristics, clinical data, and histopathologic findings, such as sex, age, primary cancer location, lymph node status, timing of first hepatectomy, number of hepatic metastases, tumor diameter, tumor distribution, preoperative serum carcinoembryonic antigen (CEA) level and (CA19-9), extent of liver resection, surgical margin, venous invasion by liver metastases, and bile duct invasion. The demographic data were collected from a prospectively maintained database. All tumors were identified routinely by contrast-enhanced computed tomography (CT) during hepatic arteriography and portography, magnetic resonance imaging (MRI), and ultrasonography (US), which were performed as preoperative studies in patients considered for the hepatectomy. The patients with resectable CRLM were treated by immediate resection or neoadjuvant chemotherapy based on the surgeon's discretion, whereas those with unresectable CRLM were treated with chemotherapy including oxaliplatin or irinotecan. Hepatectomy was planned after reassessment. It was indicated in cases where all tumors could be removed with clear margins with the remnant liver function preserved, and hepatic resection could be performed safely [16,17]. All CRLM cases were assessed for resectability by a liver surgeon, using the same criterion. During surgery, abdominal exploration and intraoperative ultrasonography were routinely used to confirm resectability, with non-anatomical liver resection with adequate free margins by the standard procedure. Major hepatectomy was defined as a resection of three or more segments as described by Couinaud, and minor hepatectomy was defined as wedge, segmental, and sectional resections. Postoperative complications were classified according to the Clavien-Dindo classification, and R0 resection was defined as no microscopic evidence of tumor at the resection margin. After initial hepatectomy, patients were routinely followed every 3 months. The levels of serum CEA and CA19-9 were measured at every outpatient clinic visit. Postoperative adjuvant chemotherapy in patients with R0 resection was not administered routinely, except in those recommended by the surgeon. Surveillance imaging, including ultrasonography and CT of the abdomen and thorax, was conducted every 3 months during the first 2 years and every 6 months until the disease-free interval exceeded 5 years.
The first relapse was defined as the site diagnosed by imaging after the initial hepatic resection, and we classified it as intrahepatic group, intra- and extrahepatic group, and extrahepatic group. The recurrent CRLM were treated with the same indication as the primary CRLM. Chemotherapy response of liver metastases was measured according to the revised response evaluation criteria in solid tumors (RECIST) criteria [18]. Categorical variables were compared using the Chi-square test, while continuous variables were compared using the ManneWhitney U-test. Overall and progression-free survival after hepatic resection were calculated by the KaplaneMeier method and compared by the log-rank test. Overall survival (OS) rates and disease-free survival (DFS) were calculated using the KaplaneMeier method. Differences between curves were assessed according to the log-rank test. Independent prognostic factors were assessed using the multivariate analysis. Cox proportional hazards model among the variables was found to be significant on univariate analysis. Differences of P < 0.05 were considered significant. All statistical analyses were performed using the JMP 10 for Windows (SAS Institute Japan). 3. Results 3.1. Comparison of clinicopathological characteristics and prognosis according to the recurrence patterns after initial hepatectomy The median follow-up for survivors as measured from the time of hepatectomy was 47.8 months. At the last follow-up, 52 patients (40.6%) had died of disease, 4 (3.1%) had died of other causes, 31 (24.2%) were alive with disease, and 41 (32.0%) were alive with no evidence of disease. Overall, 87 patients (68.0%) experienced a recurrence of disease at the last follow-up. Among the patients with recurrence, the median follow-up from the time of the initial recurrence was 27.3 months. In primary hepatectomy, 88 patients had synchronous
Fig. 2. Kaplan-Meier curves for patients without recurrence, and for those with intrahepatic, intra- and extrahepatic, and extrahepatic recurrence after initial hepatectomy, showing overall survival (A) and disease-free survival (B). Overall survival rates at 5 years were 100.0%, 46.6%, 0.0%, and 39.8%, respectively. Disease-free survival rates at 2 years were 100.0%, 6.1%, 0.0%, and 7.1%, respectively.
M. Hashimoto et al. / International Journal of Surgery 36 (2016) 96e103
metastases of CRLM and 40 had metachronous metastases of CRLM. The average number of tumors was 2.6 and average maximum size of tumors was 34.4 mm. Of the patients, 94 (73.4%) were administered adjuvant chemotherapy after primary hepatectomy and had no recurrence. After hepatectomy, 33 patients had only intrahepatic metastases, 11 had intra- and extrahepatic metastases, and 43 had only extrahepatic metastases (Fig. 1). On comparing patients in the no recurrence group and those in the intrahepatic group, those in the no recurrence group showed more well-differentiated histologic type of primary tumor than the patients from the intrahepatic group. In addition, the CEA levels and tumor numbers at the primary hepatectomy were significantly lower in the no recurrence group than in the intrahepatic group (Table 1). In contrast, no significant difference was found between the recurrence site of the intrahepatic group and the extrahepatic group as well as between the intrahepatic group and the intra- and extrahepatic group. The OS rates of the intra- and extrahepatic group were significantly lower than that of the intrahepatic group (3- and 5-year
Table 2 Overall survival rates according to clinicopathological factors. Overall survival (%)
All cases (n ¼ 33) Age, years < 65 (n ¼ 18) 65 (n ¼ 15) Gender Male (n ¼ 22) Female (n ¼ 11) Primary tumor Colon (n ¼ 20) Rectum (n ¼ 13) Histological grading Well (n ¼ 5) Moderately or poorly (n ¼ 28) Lymphatic invasion Yes (n ¼ 23) No (n ¼ 9) Venous invasion Yes (n ¼ 22) No (n ¼ 10) Depth of tumor invasion sm, mp (n ¼ 3) ss~ (n ¼ 30) Lymph node metastases Positive (n ¼ 24) Negative (n ¼ 9) Tumor size (max; mm) < 30 (n ¼ 17) 30 (n ¼ 16) Number of tumors Single (n ¼ 16) Multiple (n ¼ 17) CEA < 20 (n ¼ 19) 20 (n ¼ 14) CA19-9 < 40 (n ¼ 18) 40 (n ¼ 15) Type of hepatectomy Seggmentectomy or less (n ¼ 27) Sectionectomy or more (n ¼ 6) Disease free interval (month) < 12 (n ¼ 22) 12 (n ¼ 11) Repeat hepatectomy Yes (n ¼ 17) No (n ¼ 16)
3-year
5-year
p value
63.1
46.6
e 0.311
70.8 56.0
52.5 42.0 0.984
60.9 68.2
49.8 40.9
61.5 67.7
46.2 47.4
30.0 68.9
30.0 50.1
64.1 59.3
41.2 59.3
61.1 67.5
40.7 56.3
66.7 63.3
66.7 44.2
59.2 74.1
36.9 74.1
65.3 59.1
48.9 44.3
65.7 61.4
57.5 34.5
72.9 43.6
60.7 21.8
69.6 54.4
61.9 29.2
67.7 41.7
48.8 41.7
0.761
99
survival rates of 20.0% and 0.0%, versus 63.1% and 46.6%, respectively) (Fig. 2). In addition, the patients from the “no recurrence” group did not die. 3.2. Analysis in the intrahepatic recurrence after primary hepatectomy Table 2 summarizes the results of the univariate analyses for OS rates according to clinicopathological factors in intrahepatic recurrence after hepatectomy. Disease free interval <12 months and non-repeat hepatectomy were significant adverse prognostic factors for OS in patients with the intrahepatic recurrence after primary hepatectomy (Fig. 3). Table 3 summarizes the results of the comparison of clinicopathological characteristics between the recurrence <12 months (n ¼ 22) and the recurrence 12 months (n ¼ 11). CEA levels at the primary hepatectomy were significantly higher in the recurrence <12 months group than in the recurrence 12 months group. The percentages of patients with venous invasion of the primary tumor and CA19-9 levels at the primary hepatectomy tended to be higher in the recurrence <12 months group than in the recurrence 12 months group. Table 4 summarizes the results of the comparison of clinicopathological characteristics between patients with repeat hepatectomy (n ¼ 19) and those without repeat hepatectomy (n ¼ 14). The disease-free interval (DFI) after the primary hepatectomy was significantly shorter in patients with repeat hepatectomy than that in patients without repeat hepatectomy. 3.3. Analyses of treatment of patients with DFI <12 months after initial hepatectomy Twenty-two patients with DFI <12 months after primary
0.159
0.655
0.410
0.942
0.343
0.881
0.330
0.126
0.285
0.354
0.001 40.9 100
18.2 90.0
87.4 31.3
80.7 0.0
<0.001
Fig. 3. (A) Kaplan-Meier curves for patients with DFI < 12 months and those with DFI 12 months after initial hepatectomy showing overall survival. OS rates with DFI <12 months at 3 and 5 years (40.9% and 18.2%, respectively) were significantly lower than for those with DFI 12 months (100.0% and 90.0%, respectively) (p ¼ 0.001). (B) Kaplan-Meier curves for patients with and without repeat hepatectomy after initial hepatectomy showing overall survival. OS rates in the non-repeat hepatectomy group at 3 and 5 years (31.3% and 0.0%, respectively) were significantly lower than those for the repeat hepatectomy group (87.4% and 80.7%, respectively) (p < 0.001).
100
M. Hashimoto et al. / International Journal of Surgery 36 (2016) 96e103
Table 3 Comparison of patients characteristics with intrahepatic recurrences accoding to DFI. Variables Patients characteristics Age, years Gender Male Female DM Positive Negative Primary tumor characteristics Primary tumor site Colon Rectum T factor ~mp ss~ N factor Positive Negative Lymphatic invasion Positive Negative Venous invasion Positive Negative Histologic type well mod. por. etc Liver metastases characteristics CEA CA19-9 Preoperative chemotherapy Tumor number Tumor number Single Multiple Tumor hemi bi Maximum tumor size Surgical procedures Minor hepatectomy Major hepatectomy Blood loss Postoperative factors Complication (clavian-Dindo) III
<12 months (n ¼ 22)
12 months (n ¼ 11)
p-value
65 (40e80)
60 (50e75)
0.504 0.709
14 (63.6) 8 (36.4)
8 (72.7) 3 (27.3)
3 (13.6) 19 (86.4)
2 (18.2) 9 (81.8)
13 (59.1) 9 (40.9)
7 (63.6) 4 (36.4)
2 (9.1) 20 (90.9)
1 (9.1) 10 (90.9)
17 (77.3) 5 (22.7)
7 (63.6) 4 (36.4)
16 (76.2) 5 (23.8)
7 (63.6) 4 (36.4)
17 (81.0) 4 (19.1)
5 (45.5) 6 (54.6)
5 (22.7) 17 (77.3)
0 (0.0) 11 (100)
49.0 (0.8e436) 65.5 (1e884)
7.5 (3.1e21.7) 12 (3e53)
0.011 0.059
3 (1e13)
1 (1e5)
0.127 0.282
9 (40.9) 13 (59.1)
7 (63.6) 4 (36.4)
3 (23.1) 10 (76.9) 22 (14e60)
2 (50.0) 2 (50.0) 38 (12e50)
17 (77.3) 5 (22.7) 380 (30e1000)
10 (90.9) 1 (9.1) 190 (70e520)
5 (23.8) 16 (76.2)
3 (27.3) 8 (72.7)
27 (68.9) 10 (24.4) 4 (9.8)
25 (75.8) 5 (15.2) 3 (9.1)
7 (31.8) 15 (63.2)
10 (90.9) 1 (9.1)
1.000
1.000
1.000
0.438
0.681
0.056
0.538
0.193 0.637 0.341 1.000
0.002
Medium (range), n (%).
hepatectomy were divided into repeat hepatectomy group (n ¼ 8) and non-repeat hepatectomy group (n ¼ 14) (Fig. 4). Among the repeat hepatectomy group, 4 patients had chemotherapy before hepatectomy, whereas, 4 underwent immediate repeat hepatectomy. Table 5 shows the comparison between patients who underwent hepatectomy and those who did not among the patients with DFI <12 months. At the time of the intrahepatic recurrence, the number of intrahepatic recurrences was significantly lower in the repeat hepatectomy group than that in the non-repeat hepatectomy group. In the repeat hepatectomy group, 3 (75%) patients responded to chemotherapy with Stable Disease (SD) or better, but 1 (25%) patient had no response. However, in the non-repeat hepatectomy group, 1
(7.1%) patient had a response, but 13 (92.9%) had no response. The response to chemotherapy for intrahepatic recurrence was significantly lower in the non-repeat hepatectomy group than in the repeat hepatectomy group. 4. Discussion We investigated the patterns of recurrences that occurred after the initial hepatic resection of colorectal hepatic metastases because the recurrence after initial hepatic resection was about 70%. In addition, we particularly studied the recurrence that occurred in the liver for improving the treatment of intrahepatic recurrence after the initial hepatectomy.
M. Hashimoto et al. / International Journal of Surgery 36 (2016) 96e103
101
Table 4 Comparison of characteristics of patients with intrahepatic recurrences accoding to with or without repeat hepatectomy. Variables Patients characteristics Age, years Gender Male Female DM Positive Negative Primary tumor characteristics Primary tumor site Colon Rectum T factor ~mp ss~ N factor Positive Negative Lymphatic invasion Positive Negative Venous invasion Positive Negative Histologic type Well Mod. por. etc Liver metastases characteristics CEA CA19-9 Preoperative chemotherapy Tumor number Tumor number Single Multiple Tumor Hemi Bi Maximum tumor size Surgical procedures Minor hepatectomy Major hepatectomy Blood loss Postoperative factors Complication (clavian-Dindo) III
Repeat hepatectomy (n ¼ 17)
Non-repeat hepatectomy (n ¼ 16)
p-value
60 (40e80)
66 (45e79)
0.113 0.721
12 (70.6) 5 (29.4)
10 (62.5) 6 (37.5)
3 (17.7) 14 (82.4)
2 (12.5) 14 (87.5)
10 (58.8) 7 (41.2)
10 (62.5) 6 (37.5)
3 (17.7) 14 (82.4)
0 (0.0) 16 (100)
12 (70.6) 5 (29.4)
12 (75.0) 4 (25.0)
12 (70.6) 5 (29.4)
11 (73.3) 4 (26.7)
10 (58.8) 7 (41.2)
12 (80.0) 4 (20.0)
1 (5.9) 16 (94.1)
4 (25.0) 12 (75.0)
11.2 (2.7e235) 24 (1e884)
49.0 (0.8e436) 44 (2e505)
0.249 0.438
1 (1e13)
3 (1e13)
0.127 0.732
9 (52.9) 8 (47.1)
7 (43.8) 9 (56.3)
4 (50.0) 4 (50.0) 28 (12e60)
1 (11.1) 8 (88.9) 30 (14e50)
15 (88.2) 2 (11.8) 400 (30e1000)
12 (75.0) 4 (25.0) 380 (40e650)
3 (17.7) 14 (82.4)
4 (26.7) 11 (73.3)
14 (82.4) 3 (17.7) 0 (0.0) 14.3 (2.8e51.6)
11 (68.8) 2 (12.5) 3 (18.8) 6.3 (1.3e15.7)
1.000
1.000
0.227
1.000
1.000
0.265
0.175
0.131
0.339 0.398 0.341 0.678
1.000
0.006
Medium (range), n (%). DFI; Disease Free Interval.
Several studies reported the pattern of recurrence after initial hepatectomy in patients with CRLM [7,19]. These results were similar to those of our study. Our finding of recurrence within 2 years was also similar to that of other studie [7]. In the current study, we showed that DFI and non-repeat hepatectomy were risk factors in OS for the prognosis improvement in the intrahepatic metastases. A DFI greater than 12 months between initial hepatectomy and recurrence for CRLM has been associated with improved long-term survival in a number of studies [20e22]. In the current study, when the characteristics of the patients of DFI <12months and DFI >12months were compared, the patients of DFI <12 months had a higher serum level of CEA and underwent a nonrepeat hepatectomy. Similarly, when the characteristics of the patients in the repeat hepatectomy and without repeat hepatectomy
group were compared, those in the repeat hepatectomy group had longer DFI. A DFI <12 months was believed to be the reason for the patients' not undergoing repeat hepatectomy. Based on these results, examining the treatment choice for a better prognosis and developing a strategy to prolong the DFI is important. For hepatectomy of CRLM, the criteria for resectability of colorectal liver metastases have been significantly revised and developed recently. Timothy M. reported that the feasibility of hepatic resection should be not only be determined by the old criteria, such as the number of metastases, the size of metastases, and the extrahepatic disease, but also by R0 resection and adequate remnant liver function [23]. Moreover, repeat hepatectomy is feasible in selected patients and seems justified, as it may result in prolonged survival with acceptable rates of morbidity and
102
M. Hashimoto et al. / International Journal of Surgery 36 (2016) 96e103
Fig. 4. Kaplan-Meier curves for patients with and without repeat hepatectomy in the group with DFI <12 months after initial hepatectomy, showing overall survival. OS rates in the non-repeat hepatectomy group at 3 and 5 years (15.3% and 0.0%, respectively) were significantly lower than those for the repeat hepatectomy group (72.9% and 38.9%, respectively) (p ¼ 0.004).
mortality [22,23]. In addition, repeat hepatectomy such as 2nd, 3rd, and 4th hepatectomy results were similar to those seen after the initial hepatectomy [24]. For the recurrence after hepatectomy, the strategy of neoadjuvant chemotherapy in patients with resectable CRLM is considered because it has advantages such as an improved progression-free survival, the evaluation of chemo-responsiveness, and selection of surgery. However, for delayed surgery, cases have been reported where unresectable tumors and tumor progression occurred while treating with neo-adjuvant chemotherapy. In the current study, the same protocol could be continued in most patients; therefore, decreasing the dose of chemotherapy was necessary. As a result, only 22.2% patients had achieved the effect of the chemotherapy of Stable Disease (SD) or more. Rolf Sauer et al. proved that the toxic effect by the chemotherapy occurred more in the postoperative treatment group than in the preoperative treatment group for patients with locally advanced rectal cancer [25] and the success rates were higher in the preoperative treatment group than in the postoperative treatment group. Similarly, for the
Table 5 Patients characteristics of only colorectal liver metastases and recurrence within 12 months after hepatectomy (n ¼ 22). Variables
Repeat hepatectomy (n ¼ 8)
Non-repeat hepatectomy (n ¼ 14)
p-value
68.7 (2.7e436) 89 (1e884) 3 (1e13)
49.0 (0.8e396) 44 (2e505) 3 (1e13)
0.246 0.562 0.803 1.000
3 (37.5) 5 (62.5) 22 (15e60)
6 (42.9) 8 (57.1) 25 (14e50)
7.2 (1e87.5) 11 (1e40) 1 (1e4)
14.2 (9.6e116) 13 (2e92) 3.5 (1e7)
5 (62.5) 3 (37.5) 0 (0.0) 14.5 (10e37)
1 (7.1) 9 (64.3) 4 (28.6) 18 (10e35)
1 2 1 4
1 (7.1) 0 (0.0) 13 (92.9) 0 (0.0)
At the initial hepatectomy CEA CA19-9 Number of tumor Number of tumor Single Multiple Maximum tumor size At the intrahepatic recurrence CEA CA19-9 Number of tumor Number of tumor Single Multiple Unclear Maximum tumor size Chemotherapy after recurrence PR SD PD Non Medium (range), n (%).
(12.5) (25.0) (12.5) (50.0)
0.391 0.194 0.685 0.029 0.043
0.348 0.019
patients with CRLM, the tolerability for chemotherapy is considered to be higher in preoperative treatment group than in postoperative treatment group. In other words, controlling recurrences is thought to be difficult because of the low tolerability and inadequate amount of drug after hepatectomy. Therefore, we suggest that repeat hepatectomy is aggressively performed if the relapse tumor after initial hepatectomy is resectable. This study has several limitations to consider. First, it is a singleinstitution study and retrospective in nature. Second, the study is limited by the sample size and the long study period. Additionally, it is likely that other factors contributed to the decision for repeat hepatectomy or palliative therapy. Despite these limitations, the current results help to further clarify the characteristics of patients with recurrence following initial hepatectomy for CRLM. 5. Conclusion Patterns of recurrence after initial hepatectomy for CRLM have important implications for OS. In the intrahepatic recurrence group, short DFI was a poor prognostic factor; moreover, short DFI was correlated with CEA and CA19-9 at hepatectomy and was associated with venous invasion by the primary tumor. Ethical approval This study was reviewed and approved by the institutional review board of Hiroshima university Hospital. Sources of funding There is no source of funding for our research. Authors' contributions M. Hashimoto: study design, data collection, data analysis, drafting and revising manuscript. T. Kobayashi: study design, drafting and revising manuscript. K.Ishiyama: management of patients. K. Ide: management of patients. M. Ohira: management of patients. H. Tahara: management of patients. S. Kuroda: management of patients, data analysis; M. Hamaoka: data collection. H. Iwako: data collection. M. Okimoto: data collection. H. Ohdan: study design, drafting and revising manuscript. Conflict of interest The authors declare that they have no conflicts of interest concerning this article. Guarantor Hideki Ohdan [email protected] Department of Gastroenterological and Transplant Surgery, Applied Life Sciences, Institute of Biomedical & Health Sciences, Hiroshima University, 1-2-3 kasumi, Minami-Ku, Hiroshima 7348557, Japan. Phone numbers: þ81-082-257-5222. Acknowledgment The authors thank Dr. M Hattori (Hiroshima University, Japan) for his assistance with the statistical analysis.
M. Hashimoto et al. / International Journal of Surgery 36 (2016) 96e103
References [1] R. Siegel, D. Naishadham, A. Jemal, Cancer statistics, 2013, CA Cancer J. Clin. 63 (2013) 11e30. [2] R. Lochan, S.A. White, D.M. Manas, Liver resection for colorectal liver metastasis, Surg. Oncol. 16 (2007) 33e45. [3] J. Leporrier, J. Maurel, L. Chiche, et al., A population-based study of the incidence, management and prognosis of hepatic metastases from colorectal cancer, Br. J. Surg. 93 (2006) 465e474. [4] A. Altendorf-Hofmann, J. Scheele, A critical review of the major indicators of prognosis after resection of hepatic metastases from colorectal carcinoma, Surg. Oncol. Clin. N. Am. 12 (2003) 165e192. [5] P.C. Simmonds, J.N. Primrose, J.L. Colquitt, et al., Surgical resection of hepatic metastases from colorectal cancer: a systematic review of published studies, Br. J. Cancer 94 (2006) 982e999. [6] H. Yamada, S. Kondo, S. Okushiba, et al., Analysis of predictive factors for recurrence after hepatectomy for colorectal liver metastases, World J. Surg. 25 (2001) 1129e1133. [7] M.C. de Jong, C. Pulitano, D. Ribero, et al., Rates and patterns of recurrence following curative intent surgery for colorectal liver metastasis: an international multi-institutional analysis of 1669 patients, Ann. Surg. 250 (2009) 440e448. [8] E.K. Abdalla, J.N. Vauthey, L.M. Ellis, et al., Recurrence and outcomes following hepatic resection, radiofrequency ablation, and combined resection/ablation for colorectal liver metastases, Ann. Surg. 239 (2004) 818e825 discussion 825e817. [9] M. D'Angelica, P. Kornprat, M. Gonen, et al., Effect on outcome of recurrence patterns after hepatectomy for colorectal metastases, Ann. Surg. Oncol. 18 (2011) 1096e1103. [10] H.H. Eker, Repeat hepatectomy for recurrent colorectal metastases, Br. J. Surg. 100 (2013) 808e818. The British journal of surgery 2013; 100: 818-819. [11] L.X. Luo, Z.Y. Yu, J.W. Huang, H. Wu, Selecting patients for a second hepatectomy for colorectal metastases: an systemic review and meta-analysis, Eur. J. Surg. Oncol. J. Eur. Soc. Surg. Oncol. Br. Assoc. Surg. Oncol. 40 (2014) 1036e1048. [12] Y. Mise, K. Hasegawa, J. Shindoh, et al., The feasibility of third or more repeat
[13]
[14]
[15]
[16] [17]
[18]
[19]
[20] [21] [22]
[23] [24]
[25]
103
hepatectomy for recurrent hepatocellular carcinoma, Ann. Surg. 262 (2015) 347e357. M. Gonzalez, A. Poncet, C. Combescure, et al., Risk factors for survival after lung metastasectomy in colorectal cancer patients: a systematic review and meta-analysis, Ann. Surg. Oncol. 20 (2013) 572e579. M.G. Zampino, P. Maisonneuve, P.S. Ravenda, et al., Lung metastases from colorectal cancer: analysis of prognostic factors in a single institution study, Ann. Thorac. Surg. 98 (2014) 1238e1245. R. Adam, V. Delvart, G. Pascal, et al., Rescue surgery for unresectable colorectal liver metastases downstaged by chemotherapy: a model to predict long-term survival, Ann. Surg. 240 (2004) 644e657 discussion 657e648. M. Makuuchi, T. Kosuge, T. Takayama, et al., Surgery for small liver cancers, Seminars Surg. Oncol. 9 (1993) 298e304. T. Itamoto, K. Katayama, H. Nakahara, et al., Autologous blood storage before hepatectomy for hepatocellular carcinoma with underlying liver disease, Br. J. Surg. 90 (2003) 23e28. E.A. Eisenhauer, P. Therasse, J. Bogaerts, et al., New response evaluation criteria in solid tumours: revised RECIST guideline (version 1.1), Eur. J. cancer 45 (2009) 228e247. Y. Mise, H. Imamura, T. Hashimoto, et al., Cohort study of the survival benefit of resection for recurrent hepatic and/or pulmonary metastases after primary hepatectomy for colorectal metastases, Ann. Surg. 251 (2010) 902e909. R. Adam, G. Pascal, D. Azoulay, et al., Liver resection for colorectal metastases: the third hepatectomy, Ann. Surg. 238 (2003) 871e883 discussion 883e874. G. Sugawara, M. Isogai, Y. Kaneoka, et al., Repeat hepatectomy for recurrent colorectal metastases, Surg. Today 35 (2005) 282e289. A. Sa Cunha, C. Laurent, A. Rault, et al., A second liver resection due to recurrent colorectal liver metastases, Archives Surg. 142 (2007) 1144e1149 discussion 1150. T.M. Pawlik, R.D. Schulick, M.A. Choti, Expanding criteria for resectability of colorectal liver metastases, Oncology 13 (2008) 51e64. A. Antoniou, R.E. Lovegrove, H.S. Tilney, et al., Meta-analysis of clinical outcome after first and second liver resection for colorectal metastases, Surgery 141 (2007) 9e18. R. Sauer, H. Becker, W. Hohenberger, et al., Preoperative versus postoperative chemoradiotherapy for rectal cancer, N. Engl. J. Med. 351 (2004) 1731e1740.