Peritoneal metastasis of colorectal cancer may have a poorer prognosis than metastasis to other sites, and such metastasis is often treated palliatively.1,2,–3 Peritoneal recurrence (PR) is a rare recurrence pattern after potentially curative resection, with rates of 2–5% in previous reports.2,3,4,5,–6 Standard therapy for PR has not been established. Hyperthermic intraperitoneal chemotherapy (HIPEC) was thought to be effective for better prognosis. The multicenter phase III PRODIGE 7 trial has yet to be published but was presented in abstract form in 2018. This study did not show improved overall survival (OS) with HIPEC. In contrast, cytoreductive surgery gives satisfactory survival, and surgical resection for peritoneal metastasis also may be effective.6 There are several reports on surgical resection of synchronous peritoneal metastasis but few on PR and none related to recurrence after curative resection for PR.1,2,4,7,8

Peritoneal metastasis often is accompanied by metastases to other organs, with rates of 30–60% for PR with concurrent metastasis.2,3,–4,9 Such cases often receive systemic chemotherapy or palliative treatment. However, it has been suggested that surgical resection may lead to a better prognosis if peritoneal and hematogenous metastases are completely resected.10,11,12,–13 In our department, we perform surgery for peritoneal metastasis with and without other organ metastases. In this study, we examined the efficacy of surgical resection for PR of colorectal cancer and identified risk factors for recurrence after curative resection of PR.

Methods

Selection of Patients

Patients who underwent surgical resection for colorectal tumor at the National Cancer Center Hospital East between January 2006 and March 2014 were enrolled in the study. The exclusion criteria were patients with nonadenocarcinoma histologically and those who did not undergo radical surgery for stage 0–IV colorectal adenocarcinoma. Of the enrolled patients, those who developed PR were retrospectively reviewed, and the incidence and prognosis of PR were investigated. Clinical data were retrieved from medical records. Patient background factors, pathological findings for the primary tumor, patterns of peritoneal metastasis, and clinical outcomes were reviewed. The study protocol was approved by the National Cancer Center Institutional Review Board (2018-100). The requirement for acquisition of informed consent from patients was waived owing to the retrospective nature of this study. The study was conducted in accordance with the 1964 Declaration of Helsinki and its later amendments.

Definition, Diagnosis, and Assessment of Peritoneal Recurrence

Peritoneal recurrence included metachronous peritoneal metastasis (MPM) and peritoneal metastasis after synchronous peritoneal metastasis removal (PM-SPMR). Peritoneal recurrence was diagnosed on imaging, mainly with computed tomography (CT), and in some cases with fluorodeoxyglucose (FDG)-positron emission positron emission tomography (PET)/CT, which was performed for diagnostic support. Patients were classified using the Japanese classification of peritoneal metastasis, which is defined by the Japanese Society for Cancer of the Colon and Rectum (JSCCR) as follows: P1, metastasis localized to the peritoneum adjacent to the primary tumor; P2, metastasis limited to the peritoneum distant from the primary tumor; and P3, diffuse metastases in the peritoneum distant from the primary tumor.10 Patients also were classified into three groups based on the Peritoneal Cancer Index (PCI): < 10, 10–20, and > 20.11,12 The JSCCR classification and PCI were examined intraoperatively for cases that underwent surgery and evaluated radiologically for nonsurgical cases.

Surgical Resection of Peritoneal Recurrence

Surgical treatment was performed for metastatic lesions judged to be completely resectable, including peritoneal nodules and other organ metastases, or cases in which peritoneal nodule resection was required for symptom relief. Complete resection of peritoneal nodules macroscopically and other organ metastases (R0/1 resection) was defined as curative resection. Surgical resection of peritoneal nodules with a residual tumor macroscopically (R2 resection) was regarded as a case without curative resection.

Statistical Analysis

Categorical variables were compared by Fisher exact test. Prognostic factors for OS and risk factors for recurrence after curative resection were identified using univariate and multivariate analysis with a Cox proportional hazards regression model. Multivariate analysis was performed using a stepwise procedure with covariates with P < 0.05 in univariate analysis. Overall survival and recurrence-free survival (RFS) were estimated using the Kaplan–Meier method, and differences between survival curves were evaluated by log-rank test. P < 0.05 was considered to be significant. All statistical analyses and graphing were performed using EZR ver. 1.41 (Saitama Medical Center, Jichi Medical University, Saitama, Japan), a modified version of R commander designed for statistical functions that are frequently used in biostatistics, and a graphical user interface for R ver. 3.6.1 (The R Foundation for Statistical Computing, Vienna, Austria).13

Results

Patients Characteristics and Treatment for Peritoneal Recurrence

In our department, 2587 patients underwent surgical resection of a colorectal tumor between January 2006 and March 2014, of whom 84 with nonadenocarcinoma histologically and 247 who did not undergo radical surgery for stage 0–IV colorectal adenocarcinoma were excluded. Of the other 2256 patients, 55 of 2229 without SPM at the time of radical surgery for primary tumor developed MPM (2.5%), and 11 of 27 with SPM at the time of radical surgery for primary tumor developed PM-SPMR (40.7%). Thus, 66 cases with PR were included in the study (Fig. 1). The characteristics of these cases are shown in Table 1. The primary tumor was located in the right colon, left colon, and rectum in 24, 31, and 11 cases, respectively. The median time to PR from primary tumor resection was 16.5 months. FDG-PET/CT was performed as diagnostic support in 47 cases, of which 40 showed accumulation of FDG. The JSCCR classification and PCI were examined intraoperatively in 46 patients who underwent laparotomy or laparoscopy. In the other 20 patients, these parameters were determined by imaging. PR was in classes P1, P2, and P3 in 20, 27, and 19 cases, respectively and PCI was < 10, 10–20, and > 20 in 55, 5, and 6 cases, respectively. Of all 66 cases, 38 had peritoneum metastasis only, and 28 had concurrent metastases at other sites, including 15 cases at one site and 13 at two or more sites.

Fig. 1
figure 1

Flow chart of included and excluded patients. NET neuroendocrine tumor; SCC squamous cell carcinoma

Full size image
Table 1 Patient characteristics
Full size table

Surgical treatment was performed in 41 cases, including 18 with surgical resection only and 23 with surgical resection and chemotherapy. No patients received HIPEC. Chemotherapy after PR was administered in 42 cases, including 19 patients treated with systemic chemotherapy alone. Six patients received best supportive care. These 25 patients did not undergo surgical treatment because of multiple extensive metastases at the peritoneum or other sites that were not amenable for surgical resection, a poor general condition, or rejection of surgery by the patient.

Surgical Outcomes

Six patients underwent resection of a peritoneal nodule only, and 35 patients underwent extended surgical resection, including adjacent organs. Ten patients received chemotherapy before surgical resection of a peritoneal nodule. The median operative time was 266.5 (range 69–579) min, and median blood loss was 705 (range 10–7024) ml. Seven patients developed postoperative complications of greater than Clavien–Dindo grade III.14 However, no reoperation was required, and there was no mortality. The median postoperative hospital stay was 13 (range 5–20) days. Of the 41 patients treated with surgical resection, 32 had R0 and 6 had R1 curative resection, whereas 3 had R2 resection. Of 28 patients with concurrent metastases, 12 had curative resection, including other sites of metastases. Curative resection of peritoneal nodules and concurrent metastasis was performed simultaneously for eight of these patients and metachronously for four patients. After achieving curative resection, 17 patients received postoperative chemotherapy.

Survival

The 5-year OS rate was 43.3%, and the median survival period after diagnosis of PR was 50.2 months (Fig. 2a). In univariate analysis, time to PR, JSCCR classification, number of peritoneal nodules, PCI > 20, and curative resection were identified as possible prognostic factors. Subsequent multivariate analysis showed that curative resection was a significant independent prognostic factor (Table 2). Five-year OS was significantly better in cases with curative resection than in those without curative resection (68.7% vs. 6.3%, P < 0.001) (Fig. 2b). Age < 70 years and time to PR > 1 year were significantly more frequent in cases with curative resection, and there were no P3 and PCI > 10 cases with curative resection. Concurrent metastasis also was significantly less frequent in cases with curative resection (Table 3). However, among patients with concurrent metastasis, 5-year OS was significantly better in those with curative resection than in those without curative resection (78.7% vs. 0%, P = 0.008) (Fig. 2c).

Fig. 2
figure 2

Kaplan–Meier curves for overall survival. a All patients. b All patients classified into groups with and without curative resection. c Patients with concurrent metastasis classified into groups with and without curative resection

Full size image
Table 2 Prognostic factors for overall survival in univariate and multivariate analyses
Full size table
Table 3 Comparison between cases with and without curative resection
Full size table

Recurrence after Curative Resection

After curative resection, 28 cases developed recurrences. The site was the peritoneum only in 11 cases, hematogenous or lymphatic metastases only in 5 cases, and both sites in 12 cases. In 38 patients who underwent curative resection, the 3-year RFS rate was 21.4%, and the median RFS time was 12.8 months (Fig. 3a). In univariate and multivariate analyses, concurrent metastasis was an independent risk factor for recurrence after curative resection (Table 4). Cases with concurrent metastasis recurred more frequently within 2 years after curative resection (2-year RFS: 0% vs. 39.1%, P = 0.002) (Fig. 3b).

Fig. 3
figure 3

Kaplan–Meier curves for recurrence-free survival after curative resection. a Patients after curative resection. b Patients after curative resection classified into groups with and without concurrent metastasis

Full size image
Table 4 Risk factors for recurrence after curative resection in univariate and multivariate analyses
Full size table

Discussion

In this study, we investigated the efficacy of surgical resection of PR of colorectal cancer and identified risk factors for recurrence after curative resection of PR. Patients who underwent curative resection for limited PR had a significantly better prognosis. However, recurrence was common after curative resection, particularly for patients with concurrent metastasis. Peritoneal metastasis is a well-known poor prognostic factor in several cancers, and in colorectal cancer, Franko et al. found that the prognosis of cases with peritoneal metastasis was significantly worse than that of cases with liver or lung metastasis.9 In this study, PR was defined as MPM and PM-SPMR. The prognosis of cases with metachronous metastasis is generally found to be more favorable than for those with synchronous metastasis.15,16 However, patients with MPM have as poor a prognosis as patients with SPM.17,18,–19 PR is less commonly reported compared with lung or liver metastases, and the rate of PR ranges from 2 to 5%.2,3,4,–5 There have been few reports on PR and surgical resection of PR, but resectable PR was not uncommon in this study. To the best of our knowledge, there have been no reports on recurrence after curative resection.

Cytoreductive surgery with HIPEC contributes to improved OS of patients with peritoneal metastasis.20,21,22,–23 However, in the UNICANCER phase III trial, addition of HIPEC did not improve OS, whereas curative resection (complete or near-complete peritoneal nodule resection) gave satisfactory survival with median OS of 41.2 months and median RFS of 11.1 months.6 Furthermore, high morbidity and mortality occur after HIPEC.24 In Japanese studies, the significance of macroscopic resection of SPM without HIPEC has been shown, but macroscopic resection of PR has not been examined.1,25,26 Nagata et al. reported that surgical resection of PR had a positive effect on prognosis, but the procedure used for peritoneal nodule resection was relatively palliative.4 Our study clearly shows the prognostic significance of curative resection of PR without HIPEC.

In our department, we perform surgery for limited and resectable peritoneal metastasis. We plan curative resection if it is judged that complete resection can be achieved for a P1 or P2 recurrence, including other organ metastases. In previous reports, the surgical resection rate of PR has ranged from 9 to 34%.2,4 In this study, surgical resection was performed in 41 of 66 cases (62.1%). This higher rate is due to performance of surgery for cases requiring extended resection of adjacent organs. Curative resection was achieved in 38 of 41 cases (92.7%). Although surgery for PR seems to be highly invasive, the short-term outcomes in the postoperative course showed no complications needing reoperation and no deaths related to surgery. Consequently, curative resection led to remarkably positive outcomes in our cases.

Peritoneal metastasis often is accompanied by other organ metastases, with reported rates of concurrent metastasis with PR of 30–60%.2,3,–4,9 Concurrent metastasis also has been reported to be a poor prognostic factor.4 In the present study, 28 cases (42.4%) had PR and simultaneous metastasis at another site. Such cases often receive systemic chemotherapy or palliative treatment, but the efficacy of surgical resection for cases with peritoneal and liver metastases has been shown.27,28,29,–30 In this study, five cases with peritoneal and liver metastases had a good prognosis after curative resection. Therefore, surgery should be considered if curative resection is possible for peritoneal metastasis and metastases at other sites.

There are no previous reports on recurrence after curative resection of PR from colorectal cancer. We found a recurrence rate of > 70% at 3 years after curative resection, which is particularly high. Despite the small number of cases, concurrent metastasis was identified as an independent risk factor, and most cases with this factor developed recurrence within 2 years. The recurrence site after curative resection was the peritoneum only in 11 cases, hematogenous or lymphatic metastases only in 12 cases, and both sites in 5 cases. The median RFS was about 12 months. Second-look surgery has been proposed for high risk patients with PR after resection of a primary tumor, and this strategy may be considered about 1 year after curative resection of PR.31,32 There is no standard adjuvant therapy after curative resection of PR. In this study, 17 of 38 patients (44.7%) underwent chemotherapy after curative resection. The high number of recurrences outside the peritoneal site suggests that there may be merit to adjuvant chemotherapy after surgery.

This study has several limitations. First, it was a retrospective review at a single center. Second, although the number of radical surgeries for stage 0-IV colorectal adenocarcinoma was a sufficient sample size compared with previous reports, the number of PR cases was relatively small. Third, the PCI status without surgery seemed to be milder than in previous reports, which may be because PCI scores were evaluated radiologically and might be underestimated.33 Also, PR might have been diagnosed at a relatively early stage, because we performed FDG-PET/CT for many patients. Fourth, most cases of PR in this study were solitary nodules or had a low PCI < 10. There is no method for distinguishing a solitary peritoneal nodule close to the primary tumor location (P1) from local recurrence. However, a solitary nodule located far from the primary tumor location (P2) should be regarded as PR. Fifth, standard chemotherapy regimens changed over the long study period. Therefore, chemotherapy for the subjects included various regimens.

Conclusions

Curative resection improved overall survival in patients with PR of colorectal cancer with or without concurrent metastasis. This treatment approach should be considered for cases with limited and resectable PR, even with concurrent metastasis. However, the recurrence rate after curative resection of PR was extremely high. Therefore, additional management and treatment for PR cases are needed, especially in those with concurrent metastasis.