Ann Surg Oncol DOI 10.1245/s10434-017-6133-7

ORIGINAL ARTICLE – COLORECTAL CANCER

Long-Term Outcomes After R0 Resection of Synchronous Peritoneal Metastasis from Colorectal Cancer Without Cytoreductive Surgery or Hyperthermic Intraperitoneal Chemotherapy Dai Shida, MD, PhD, Shunsuke Tsukamoto, MD, PhD, Hiroki Ochiai, MD, PhD, and Yukihide Kanemitsu, MD Colorectal Surgery Division, National Cancer Center Hospital, Tokyo, Japan

ABSTRACT Background. The National Comprehensive Cancer Network (NCCN) guidelines for colon cancer recently added the following footnote regarding the therapeutic strategy for peritoneal metastases: ‘‘If R0 resection can be achieved, surgical resection of isolated peritoneal disease may be considered at experienced centers.’’ This study investigated the efficacy of R0 resection of peritoneal metastasis from colorectal cancer without cytoreductive surgery or hyperthermic intraperitoneal chemotherapy (HIPEC). Methods. This retrospective cohort study was conducted at a single-institution tertiary care cancer center. Among 496 consecutive M1c colorectal cancer patients, R0 resection was achieved for 94 patients (19%). The subjects were 78 consecutive patients with colorectal cancer and simultaneous peritoneal metastasis but no other distant metastases who underwent R0 resection at the National Cancer Center Hospital from 1971 to 2016 (16% of all M1c patients). Overall survival (OS) was investigated, and clinicopathologic variables were analyzed for prognostic significance. Results. No perioperative mortality was noted. The 3-year OS rate was 45%, and the 5-year OS rate was 28.7%. The median survival time was 33.4 months. Notably, 17 patients survived for more than 5 years, and 9 of these patients did not receive any chemotherapy. Multivariate analysis showed cancer location in the colon and

Ó Society of Surgical Oncology 2017 First Received: 5 August 2017 D. Shida, MD, PhD e-mail: [email protected]

harvesting of 12 or more lymph nodes to be independent factors associated with a better prognosis. Conclusions. From the perspective of long-term outcomes and no perioperative mortality, R0 resection of peritoneal metastasis from colorectal cancer, without complete peritonectomy or HIPEC, appeared to be an acceptable therapeutic option for some patients with peritoneal metastasis. Keywords Peritoneal metastasis
Colorectal cancer
R0 resection
NCCN guidelines The peritoneum is a common site for metastasis in advanced-stage colorectal cancer patients.1 Given the poor prognosis for peritoneal metastasis relative to metastatic disease in visceral organs,2 its tumor-node-metastasis (TNM) classification (8th ed published in 2017) was revised to M1c (metastasis to the peritoneum with or without other organ involvement) and separated from M1a (metastasis to one organ) and M1b (metastasis to more than one organ).3 Moreover, the stage classification of peritoneal metastasis was revised to stage 4C.3 Aggressive cytoreductive debulking surgery (peritonealstripping surgery) in combination with hyperthermic intraperitoneal chemotherapy (HIPEC) is a treatment option for peritoneal metastasis.4 However, given the significant morbidity and mortality associated with this procedure and the lack of evidence supporting its benefits, the National Comprehensive Cancer Network (NCCN) guidelines currently consider this treatment to be nonstandard and do not endorse its use outside clinical trials.5 Thus, most oncologists view colorectal cancer with peritoneal metastasis as a terminal condition only to be palliated with systemic chemotherapy and have traditionally considered surgical treatment to be futile.

D. Shida et al.

Against this backdrop, the NCCN guidelines for colon cancer were recently revised (2016 version 1) to add a footnote regarding the therapeutic strategy for peritoneal metastasis (page COL-8) as follows: ‘‘Aggressive cytoreductive debulking and/or intraperitoneal chemotherapy are not recommended outside the setting of a clinical trial. If R0 resection can be achieved, surgical resection of isolated peritoneal disease may be considered at experienced centers.’’5 The R0 resection of peritoneal metastasis, which differs from aggressive cytoreductive surgery, is a new concept in the NCCN guidelines. With this revision, the efficacy of R0 resection of peritoneal metastasis for local control is currently drawing attention. In this study, we hypothesized that peritoneal metastasis is not always a terminal condition only to be palliated with systemic chemotherapy, but may represent a first site of metastasis in some cases and may not necessarily be indicative of general disease. Accordingly, the current study aimed to analyze long-term outcomes after R0 resection of synchronous peritoneal metastasis from colorectal cancer, as described by the revised NCCN guidelines, in order to assess whether surgical intervention has any clinical benefit. We also sought to identify prognostic factors that may aid in the development of appropriate operative strategies. METHODS Patient Selection and Data The current study was retrospective in design. The inclusion criteria for the study specified patients with synchronous peritoneal metastasis of colorectal adenocarcinoma who were referred to the National Cancer Center Hospital and underwent R0 resection from January 1971 to December 2016. All tumors of peritoneal metastasis were resected and histologically proven to be peritoneal metastasis of colorectal cancer. The final study population consisted of 78 patients who met the inclusion criteria of R0 resection for M1c without distant metastasis other than to the peritoneum (Fig. 1). The following parameters were assessed: age, gender, operation year, serum levels of carcinoembryonic antigen (CEA), location of primary tumor, histologic type, tumor size, depth of tumor invasion, and lymph node metastasis. The Union for International Cancer Control (UICC) TNM classification (8th ed) was adopted.3 The extent of peritoneal metastasis was defined as metastasis only to the adjacent peritoneum (P1), a few metastases to the distant peritoneum (P2), or diffuse metastases to the distant peritoneum (P3), in accordance with the Japanese classification of colorectal carcinoma as defined by the Japan Society for

Operation for Stage IV colorectal cancer (1971-2016) (n=10 858) Synchronous peritoneal metastasis (n=496) 4.6% of all stage IV patients

R0 resection (n=94) for M1c Including other distant metastases 19% of all M1c patients

Concomitant with other distant metastases (n=16)

R0 resection for M1c without other distant metastases (n=78) 16% of all M1c patients

FIG. 1 Current study cohort. After exclusion of patients with cancers involving other distant metastases (n = 16) from the initial 94 patients who underwent R0 resection for M1c, the final study population consisted of 78 patients

Cancer of the Colon and Rectum (JSCCR).6 According to the Study Group for Peritoneal Metastasis From Colorectal Cancer by the Japanese Society for Cancer of the Colon and Rectum (JSCCR), in categorizing the number of peritoneal metastases, many Japanese surgeons have classified fewer than 4 metastases as P1, 4 to 10 metastases as P2, and more than 10 metastases as P3.7 In other words, P2 also is defined as 10 or fewer peritoneal metastases disseminated in two or more areas, and P3 is defined as 10 peritoneal metastases disseminated in more than three areas. Thus, a PCI score of P1 seems to range from one to six. A PCI score of P2 seems to range from 4 to 20, and a PCI score of P3 seems to be higher than 10. In addition to 5-fluorouracil (5-FU), an oxaliplatincontaining regimen has been used as adjuvant chemotherapy since 2009, when oxaliplatin began to be covered by health insurance in Japan. This study was approved by the Institutional Review Board (IRB) of the National Cancer Center Hospital (IRB code 2015-320). R0 Resection of Peritoneal Metastasis from Colorectal Cancer Without Cytoreductive Surgery or HIPEC Before tumor resection, we examined the entire abdominal region by visual inspection and palpation, including the local region, Douglas’ pouch, and liver surface, because peritoneal metastasis from colorectal cancer, especially small peritoneal metastases, are difficult to detect preoperatively and often are discovered during initial surgery. Staging laparoscopy or staging laparotomy was never used for this purpose.

R0 Resection of Peritoneal Metastasis

After examination of the entire abdomen, if peritoneal metastasis existed as P1 or P2, for which R0 resection could be achieved, all peritoneal metastases that could be detected macroscopically were dissected with resection of the primary tumor together with regional lymph node dissection in the same initial operation such that no macroscopic tumors remained. In the current cases, only the diseased portion of the peritoneum was dissected. This differed from traditional cytoreductive surgery, which resects not only the diseased portion, but also the entire peritoneum (peritonectomy). Whereas cytoreductive surgery for colorectal peritoneal disease recently does not necessarily mandate stripping of the entire peritoneum, at least the NCCN guidelines (2016 version) clearly distinguish R0 resection from cytoreductive surgery. Among the 78 patients in this study, resection of the ovary was required for 8 patients, greater omentum for 6 patients, small bowel for 4 patients, bladder for 5 patients, and uterus for 2 patients. We did not use intraoperative chemotherapy, including HIPEC. Follow-up Evaluation Postoperative follow-up evaluation consisted of computed tomography (CT) scan and serum tumor marker measurements every 3 months for the first 2 years, then every 6 months for 5 years. Follow-up data were documented prospectively until an event or until the study cutoff date of May 2017. Complete follow-up assessment was performed for the entire cohort of patients, with a median follow-up time of 49 months (range, 2–244 months) for survivors. Statistical Analysis Pearson’s v2 test for categorical variables and the Wilcoxon rank-sum test for continuous variables were used to examine various factors. The primary study end point was overall survival (OS), or all-cause mortality, defined as the interval between the date of operation and the date of either death or the end of the observation period. Patients alive at the end of follow-up period were censored. The Kaplan–Meier method was used to estimate OS. Differences in survival were assessed with the log-rank test. Multivariate Cox proportional hazards regression models were subsequently fitted to evaluate the relationship between R0 resection of peritoneal metastasis from colorectal cancer and OS, with control used for potential confounding covariates. Independent prognostic factors for colorectal cancer patients with peritoneal metastasis were determined using Cox proportional hazards models. Data are expressed as numbers of patients, ratios (%), or hazard ratios (HR) and 95% confidence intervals (CI), as

indicated. A p value lower than 0.05 was considered statistically significant. All statistical analyses were performed using the JMP12 software program (SAS Institute Japan Ltd., Tokyo, Japan). RESULTS Characteristics of the Study Cohort The details of our study cohort are summarized in Fig. 1. As discussed earlier, patients with synchronous peritoneal metastasis who were referred to the National Cancer Center Hospital and underwent R0 resection from January 1971 to December 2016 met the inclusion criteria for this study. During the study period, 10,858 patients with stage 4 colorectal cancer underwent some type of operation, such as tumor resection, diverting stoma construction, bypass surgery, or probe laparotomy. Of these patients, 496 (4.6% of all the stage 4 patients) had synchronous peritoneal metastasis (M1c and stage 4C). Among these patients, 94 underwent R0 resection. Because our analysis focused on proving the hypothesis that peritoneal metastasis may represent the first site of metastasis in some cases, we excluded 16 patients with distant metastases to organs other than the peritoneum, such as the liver (n = 11) and non-regional lymph nodes (n = 5), considering that colorectal cancer likely had already spread systematically in these patients. Accordingly, the final study population consisted of 78 consecutive patients (16% of all the M1c/stage 4C patients) with colorectal cancer and simultaneous peritoneal metastasis, without other distant metastases, who underwent R0 resection (Fig. 1). The patient characteristics are shown in Table 1. The median patient age was 61 years (range, 26–91 years). Among the 78 patients, 64 had a primary tumor in the colon, and 14 had a primary tumor in the rectum. Regarding the extent of peritoneal metastasis, 48 patients had P1, and 30 patients had P2.6 No patients had diffuse metastases in the distant peritoneum (P3). Of the 78 patients, 75 underwent open surgery, and 3 underwent laparoscopic surgery. Adjuvant chemotherapy with 5-FU was administered after R0 surgery to 39 patients (50%), 11 (14%) of whom received intensive chemotherapy containing oxaliplatin. None of the patients in our series received HIPEC. Long-Term Outcomes After R0 Surgery The study had no cases of perioperative mortality. Regarding long-term outcomes, the 3-year OS rate was

D. Shida et al. TABLE 1 Uni- and multivariate analyses of factors affecting survival among colorectal cancer patients with peritoneal metastases after R0 resection Variable

Operation year Gender Age (years) Preoperative CEA level

Category

N

Median overall survival (months)

Univariate analysis (p value) 0.0936

1971–1990

32

28.6 (20.6–42.5)

1991–2016

46

46.3 (28.3–67.7)

Male

38

28.3 (20.8–59.2)

Female

40

42.5 (28.8–51.2)

\ 70

57

33.5 (25.0–43.6)

C 70

21

49.6 (23.7–82.5)

\ 30

61

31.4 (25.1–50.1)

C 30

10

43.6 (17.4–)

Unknown

7

P1

48

30.3 (25.3–49.6)

P2

30

42.5 (19.0–57.1)

Colon

64

42.7 (28.8–57.1)

Rectum

14

21.1 (11.0–36.0)

Size of primary tumor (mm)

\ 70 C 70

60 18

Histology

Differentiated Others

P classification Tumor location

Depth of invasion No. of dissected LN

HR

95% CI

p Value

0.7506 0.6732 0.4335

0.7212 0.0124

1 2.12

1.06–3.98

0.0352

36.0 (28.8–53.2) 23.1 (16.5–42.5)

0.0921

1 1.53

0.75–2.98

0.2351

51

46.7 (28.3–64.8)

0.033

1

27

25.2 (17.4–36.0)

0.64–2.41

0.4943

1.01–4.44

0.0462

T2/T3

15

43.6 (23.0–)

T4

63

30.0 (23.7–46.3)

1.26 0.246

C 12

65

42.7 (28.3–53.2)

\ 12

13

22.8 (8.9–31.4)

LN metastasis

N0

14

36.0 (18.3–53.2)

N1/N2

64

31.4 (25.2–50.1)

Postoperative chemotherapy \1[

Absent

39

33.9 (23.1–53.2)

Present

39

33.5 (25.0–51.2)

Postoperative chemotherapy \2[

Multivariate analysis

None/5FU only

67

31.4 (25.0–43.6)

5FU ? OX

11

51.2 (23.4–)

0.0181

1 2.19

0.8916 0.992 0.1685

HR hazard ratio, CI confidence interval, CEA carcinoembryonic antigen, LN lymph node, OX oxaliplatin

1.0

Overall survival

45%, and the 5-year OS rate was 28.7%, with a median follow-up time of 49 months for survivors. The median survival time was 33.4 months. Notably, 17 patients survived longer than 5 years (Fig. 2). During the study period, recurrence after R0 surgery was observed in 47 patients (60.3%), with the peritoneum as the most frequent recurrence site (n = 26). Hematogenous recurrence in the liver (n = 14), lungs (n = 7), bone (n = 1), and brain (n = 1) as well as local recurrence (n = 5) also were observed. Of the 78 patients, 39 (50%) received adjuvant chemotherapy, 28 of whom were treated with 5-FU monotherapy. The survival curves for the groups with and without adjuvant chemotherapy were similar, with no significant difference between the groups (p = 0.992) (data not shown).

0.8 0.6 33.4 months

0.4

45.0%

R0 resection (n= 78) 28.7%

0.2 0.0 1

5

3

Years after surgery No. at risk

78

66

48

32

25

17

15

FIG. 2 Overall survival curve of patients who underwent R0 resection for synchronous peritoneal metastasis without other distant metastases (n = 78)

R0 Resection of Peritoneal Metastasis

Factors Affecting Prognosis After R0 Surgery According to the univariate analysis, location of primary cancer in the colon (p = 0.012), differentiated histologic type (p = 0.033), and harvesting of 12 or more lymph nodes (p = 0.018) were significantly associated with better OS, whereas operation year, gender, age, and preoperative carcinoembryonic antigen (CEA)0 level were not associated with prognosis (Table 1). Although not significant, patients with tumors 7 cm or larger tended to have a better prognosis (p = 0.092). The multivariate Cox proportional hazards regression models showed that of the four prognostic factors, location of primary cancer in the colon (HR of the rectum, 2.12; 95% CI, 1.06–3.98; p = 0.035) and harvesting of 12 or more lymph nodes (HR of harvesting fewer than 12 lymph nodes, 2.19; 95% CI, 1.01–4.44; p = 0.046) were independent factors associated with a better prognosis (Table 1). Long-Term Survival After R0 Surgery A total of 17 patients survived for more than 5 years after R0 surgery. In 16 of these patients, the primary tumor was located in the colon. Regarding the extent of peritoneal metastasis, 12 patients were P1, and 5 patients were P2. Eight patients received chemotherapy after surgery as adjuvant chemotherapy (7 received 5-FU and 1 received 5-FU plus oxaliplatin). None of the patients received chemotherapy before or during surgery. Nine patients were treated by surgery alone and did not receive any chemotherapy during the perioperative period. Only one of these nine patients received chemotherapy after the perioperative period because of metachronous metastasis to the lung, which was later resected. DISCUSSION This study found that the median survival time for the patients with R0 resection of synchronous peritoneal metastasis from colorectal cancer, without cytoreductive surgery or HIPEC, was 33.4 months and that the 5-year OS rate was 28.7%. The median 5-year survival in previous studies for patients undergoing cytoreductive surgery with HIPEC was reported to be an average of 30 months for those with limited peritoneal metastases, and one paper showed the 5-year OS rate as 29.7% for colon cancer and 37.9% for rectal cancer.8–10 Even patients who underwent resection of liver metastases, who are expected to have a favorable prognosis among stage 4 colorectal cancer patients, had 5-year survival rates of about 40%.11,12 Taken together, from the perspective of long-term outcomes, the strategy of R0 resection of synchronous peritoneal metastasis from colorectal cancer appears to be an

acceptable option. Harvesting of 12 or more lymph nodes was found to be an independent factor associated with better prognosis. Thus, when R0 resection appears to be achievable for synchronous peritoneal metastasis from colorectal cancer, colectomy with sufficient local lymph node dissection, which could result in the harvesting of 12 or more lymph nodes, and resection of visible peritoneal metastases are recommended. Japan has unique therapeutic strategies for addressing peritoneal metastasis of colorectal cancer, which differ from those of Western countries.1 Specifically, complete resection is desirable for P1, and complete resection is considered for P2 when it is easily resectable.1 A 2014 publication using a database from a nationwide multicenter registry of the JSCCR that covers approximately 10% of all patients with colorectal cancer in Japan reported that none of the M1c patients received HIPEC.13,14 A study group of the JSCCR found that in a multi-institutional retrospective analysis, the overall 5-year survival rate was 36% for 72 patients with P1 and no other distant metastases who underwent R0 surgery.7,15–18 However, in their series, due to insufficient background records and insufficient follow-up information, only one third of R0 resection cases could be analyzed (72 of 224 patients). Thus, the results may not have been representative of the entire study population.15 Notably, 17 patients in the current study survived longer than 5 years, and 9 of them were treated by surgery alone and did not receive any chemotherapy during the perioperative period. This means R0 resection without cytoreductive surgery and HIPEC actually could control peritoneal disease for these patients. This suggests that peritoneal carcinomatosis is not always a terminal condition and that in some cases, peritoneal metastasis of colorectal cancer initially remains local before becoming a systemic disease. In such cases, surgical resection of peritoneal disease as local control, either by R0 resection or by cytoreductive surgery, is important. On the other hand, considering that recurrence sites after R0 resection for colorectal cancer include not only the peritoneum, but also the liver and lungs, peritoneal metastasis likely progresses to systemic disease at some point. The R0 resection rate among all the patients with synchronous peritoneal metastasis (M1c) in the current study was 19%, which remained low due to diffuse peritoneal metastasis, synchronous hematogenous metastasis, or both. The remainder of the patients (i.e., 81% of the M1c patients) likely required chemotherapy to prolong survival. Despite advances in cytotoxic chemotherapy, the median survival time for patients with metastatic colorectal cancer is less than 3 years, and patients with peritoneal metastatic colorectal cancer have a significantly shorter OS time than those with metastases to other isolated sites.2 The most

D. Shida et al.

appropriate regimen for peritoneal metastasis of colorectal cancer will need to be clarified in future studies. This study had some potential limitations. First, because the current study was retrospective in design, biases may have existed. Second, the sample size was relatively small, although the number of R0 patients with sufficient background records represented the largest sample reported to date. Third, this study did not include a control group for comparison. Fourth, although consecutive patients were enrolled, the study period was from 1971 to 2016. During that long period, treatment strategies including chemotherapy as well as the perioperative awareness of peritoneal metastasis have changed significantly. Thus, the reported data are hardly comparable with those of other studies using novel approaches. This heterogeneity made it difficult to investigate the efficacy of adjuvant chemotherapy. Nonetheless, our observations warrant further consideration and validation in a larger patient series of colorectal cancer with synchronous peritoneal metastases. In conclusion, from the perspective of long-term outcomes and no perioperative mortality, R0 resection of peritoneal metastasis from colorectal cancer, without performance of entire peritonectomy or HIPEC, appeared to be acceptable as a treatment option. In some cases, peritoneal metastasis of colorectal cancer remains local before it progresses to systemic disease. In these cases, surgical resection of isolated peritoneal disease with sufficient local lymph node dissection is desirable.

5.

6. 7.

8.

9.

10.

11.

12.

13.

14. DISCLOSURES

There are no conflicts of interest. 15.

REFERENCES 1. Watanabe T, Muro K, Ajioka Y, et al. Japanese Society for Cancer of the Colon and Rectum (JSCCR) guidelines 2016 for the treatment of colorectal cancer. Int J Clin Oncol. 2017. doi:10. 1007/s10147-017-1101-6. 2. Franko J, Shi Q, Meyers JP, et al. Prognosis of patients with peritoneal metastatic colorectal cancer given systemic therapy: an analysis of individual patient data from prospective randomised trials from the Analysis and Research in Cancers of the Digestive System (ARCAD) database. Lancet Oncol. 2016;17:1709–19. 3. UICC. TNM Classification of Malignant Tumours, 8th ed. Wiley: New York; 2017. 4. Sugarbaker PH, Ryan DP. Cytoreductive surgery plus hyperthermic perioperative chemotherapy to treat peritoneal metastases

16.

17.

18.

from colorectal cancer: standard of care or an experimental approach? Lancet Oncol. 2012;13:e362–9. Benson AB III, Venook AP, Cederquist L, et al. Colon Cancer, version 1.2017, NCCN Clinical Practice Guidelines in Oncology. J Natl Compr Canc Netw. 2017;15:370–98. Rectum JSfCotCa. Japanese Classification of Colorectal Carcinoma (8th ed). Kanehara & Co., Ltd: Tokyo, 2013. Kobayashi H, Kotake K, Sugihara K; Study Group for Peritoneal Metastasis from Colorectal Cancer by the Japanese Society for Cancer of the C, Rectum. Enhancing the objectivity of the Japanese classification of peritoneal metastases from colorectal cancer. Jpn J Clin Oncol. 2014;44:898–902. Elias D, Gilly F, Boutitie F, et al. Peritoneal colorectal carcinomatosis treated with surgery and perioperative intraperitoneal chemotherapy: retrospective analysis of 523 patients from a multicentric French study. J Clin Oncol. 2010;28:63–8. Ihemelandu C, Sugarbaker PH. Management for peritoneal metastasis of colonic origin: role of cytoreductive surgery and perioperative intraperitoneal chemotherapy: a single institution’s experience during two decades. Ann Surg Oncol. 2017;244:898–905. Elias D, Glehen O, Pocard M, et al. A comparative study of complete cytoreductive surgery plus intraperitoneal chemotherapy to treat peritoneal dissemination from colon, rectum, small bowel, and nonpseudomyxoma appendix. Ann Surg. 2010;251:896–901. Morris EJ, Forman D, Thomas JD, et al. Surgical management and outcomes of colorectal cancer liver metastases. Br J Surg. 2010;97:1110–18. Jones RP, Kokudo N, Folprecht G, et al. Colorectal liver metastases: a critical review of state of the art. Liver Cancer. 2016;6:66–71. Noura S, Ohue M, Ito Y, et al. New staging system for colorectal cancer patients with synchronous peritoneal metastasis in accordance with the Japanese classification of colorectal carcinoma: a multi-institutional study. Dig Surg. 2016;33:66–73. Kobayashi H, Kotake K, Sugihara K. Outcomes of surgery without HIPEC for synchronous peritoneal metastasis from colorectal cancer: data from a multi-center registry. Int J Clin Oncol. 2014;19:98–105. Sato H, Maeda K, Kotake K, Sugihara K, Takahashi H. Factors affecting recurrence and prognosis after R0 resection for colorectal cancer with peritoneal metastasis. J Gastroenterol. 2016;51:465–72. Ozawa H, Kotake K, Kobayashi H, Kobayashi H, Sugihara K. Prognostic factors for peritoneal carcinomatosis originating from colorectal cancer: an analysis of 921 patients from a multi-institutional database. Surg Today. 2014;44:1643–50. Kobayashi H, Kotake K, Funahashi K, et al. Clinical benefit of surgery for stage IV colorectal cancer with synchronous peritoneal metastasis. J Gastroenterol. 2014;49:646–54. Furuhata T, Okita K, Nishidate T, et al. Oncological benefit of primary tumor resection with high tie lymph node dissection in unresectable colorectal cancer with synchronous peritoneal metastasis: a propensity score analysis of data from a multi-institute database. Int J Clin Oncol. 2015;20:922–7.