Long Term Survival after Cytoreductive Surgery Combined with Perioperative Chemotherapy in Gastric Cancer Patients with Peritoneal Metastasis

The present study demonstrated prognostic factors for long-term survival in patients after a comprehensive treatment (CHT) for peritoneal metastasis (PM) from gastric cancer (GC). Materials and Methods: Among 419 patients treated with neoadjuvant intraperitoneal/systemic chemotherapy (NIPS), 266 (63.5%) patients received complete resection (CC-0) of the macroscopic tumors. In total, 184 (43.9%) patients were treated with postoperative systemic chemotherapy. Results: All patients treated who received incomplete cytoreduction (CC-1) died of GC within 6 years. In contrast, 10- year survival rates (-YSR) of CC-0 resection were 8.3% with median survival time (MST) of 20.5 months. Post-NIPS peritoneal cancer index (PCI) ≤11, and pre-NIPS PCI ≤13 were the significant favorable prognostic factors. Patients with numbers of involved peritoneal sectors ≤5 survived significant longer than those with ≥6. Both negative pre- and post-NIPS cytology was associated with significant favorable prognosis. Multivariate analyses identified pre-PCI (≤13 vs. ≥14), and cytology after NIPS (negative cytology vs. positive cytology) as independent prognostic factors. Ten year-survivors were found in patients with involvement of the greater omentum (9%), pelvic peritoneum (3%), para-colic gutter (13.9%), upper jejunum (5.6%), lower jejunum (5.5%), spermatic cord (21.9%), rectum (9.5%), ureter (6.3%), ovary (6.7%), and diaphragm (7.0%) at the time of cytoreduction. Twenty-one patients survived longer than 5 years, and 17 patients are still alive without recurrence. Conclusions: GC-PM should be removed aggressively, in patients with PCI after NIPS ≤11, PCI before NIPS ≤13, mall bowel PCI ≤2, and complete cytoreduction should be performed for metastasis in ≤5 peritoneal sectors.


Introduction
Despite the recent development of new chemotherapeutic drugs and molecular targeted drugs, the results of systemic chemotherapy for peritoneal metastasis (PM) from gastric cancer (GC) patients remain very poor [1][2][3]. Hong et al. reported that median overall survival of GC patients with PM treated with systemic chemotherapy alone was 12.5 months (95% CI; 9.4-15.5 months) and all patients died within 5 years [4].
In 1999, the Peritoneal Surface Oncology Group International (PSOGI) proposed a novel therapeutic approach, combining cytoreductive surgery and perioperative chemotherapy (POC) for patients with PM [5]. The treatment is a comprehensive treatment, and its purpose is to cure patients with PM by use of a combination of complete resection of visible metastasis and POC to eradicate micrometastatic disease. Yonemura and Glehen reported that the overall median survival time (MST) after the treatment ranged from 9.5 to 20.5 months, and 5-year survival rates were 18% and 13% [6,7]. Long term survival was significantly better after the treatment than after systemic chemotherapy alone [4,6,7].
So far, however, no study has reported 10-year survival, and the survivals of patients with special reference to the spread of PMs and the metastatic involvement of peritoneal sectors and organs after the comprehensive treatment. The present study analyzed long-term survival in patients with GC-PM after comprehensive treatment to identify prognostic factors affecting survival and to demonstrate the efficacy of the removal of the peritoneum and organs infiltrated by PM. The aim of this study was to evaluate long-term survival according to the different peritoneal sectors and organs removed.

Materials and Methods
This retrospective observational study, included in total 419 patients with GC-PM, who were treated with preoperative intraperitoneal/systemic chemotherapy (NIPS), cytoreductive surgery (CRS), and hyperthermic intraperitoneal chemoperfusion (HIPEC) between 2006 and 2019 at our institute. Inclusion criteria were pathologically proven adenocarcinoma of the stomach and pathologically proven peritoneal metastasis or positive cytology. All patients were treated with NIPS. The number of primary and recurrent cases were 281, and 138, respectively A peritoneal port system (Hickman Subcutaneous port; BARD, Salt Lake City, UT, USA) was introduced into the abdominal cavity, and the port tip was placed in the cul-de-sac. A series of 3-week cycles of NIPS was performed. Specifically, S1 was administered orally twice daily at a dose of 60 mg/m 2 /d for 14 consecutive days, followed by 7 days of rest. Docetaxel and cisplatin were administered intraperitoneally (i.p.) at a dose of 30 mg/m 2 each on day 1. Docetaxel and cisplatin were diluted in 500 mL of normal saline and administered through the peritoneal port system. The same doses of docetaxel and cisplatin were administered intravenously (IV) on day 8 after standard premedication. The treatment course was repeated every 3 weeks for 3 courses [8]. Before NIPS, 164 patients underwent exploratory laparoscopy, and peritoneal cancer index (PCI) was calculated and peritoneal cytology was studied.
Four to six weeks after the last cycle of preoperative chemotherapy, patients received a laparotomy beginning with incision from the xiphoid process to the pubic bone. After extensive intraperitoneal lavage using 10 L of saline, the PCI of each patient was scored [9]. Small bowel PCI (SB-PCI) is an index of total lesion size in 4 parts of the small bowel and its mesentery; it ranges from 0 to 12. Then standard peritonectomy procedures and visceral resection were performed according to the extent of disease from an oncological point of view [4]. All procedures were performed with the goal of obtaining completeness of cytoreduction (CC)-0 resection [10,11]. Following complete CRS, 255 patients were treated with HIPEC with mitomycin C (MMC) at a dose of 12.5 mg/m 2 and cisplatin (CDDP) at 50 mg/m 2 in 4 L of saline at an intraperitoneal temperature between 42 • C and 43.5 • C.
Removed organs and peritoneum were histologically studied by hematoxylin-eosin staining. The pathological changes related to growth of the primary tumors and peritoneal lesions were observed by two pathologists. The basic pathological changes were classified according to Japanese Classification of Gastric Carcinoma [12]. The grade ranged from 0 to 3, with 0 indicating no evidence of effect; 1a, viable tumor cells occupy more than 2/3 of the tumor area; 1b, viable tumor cells occupy more than 1/3 but less than 2/3 of the tumor area; 2, viable tumor cells remain in less than 1/3 of the tumor area; grade 3, no viable tumor cells remain. In the study, grade 0 and 1a were considered histological non-responders, and grade 1b, 2, and 3 were considered as responders. All tumors were macroscopically classified according to the Bormann classification.
After discharge, patients were followed up every 2 months until 3 years and every 6 months after 4-years as long as possible.
During follow up at the outpatient clinic, 184 patients were treated with postoperative systemic chemotherapy (40 regimens).

Ethical Standards
All subjects gave their informed consent for inclusion before they participated in the study. The study was conducted in accordance with the Declaration of Helsinki, and the protocol was approved by the Ethics Committee of Kishiwada Tokushukai and Kusatsu General Hospital ("A study of the safety and efficacy of a comprehensive treatment for the treatment of peritoneal metastasis from gastrointestinal cancer, H-19: Project identification code, on 26 October 2012"). All patients were informed about the adverse events of the procedure and gave their written informed consents to participate.

Evaluation of Complications
Postoperative complications were graded according to the system of classification reported by Dindo and colleagues [13].

Data Analysis
The survival analysis was performed using the Kaplan-Meier method and compared using the log rank test. Categorical variables were compared by X 2 analysis or Fischer's exact test. For multivariate analysis, a Cox regression model using factors which are significantly associated with death by univariate analyses was performed. Statistical analyses were performed using SPSS version 11.5 (SPSS Inc., Chicago, IL, USA). Confidence interval were calculated at the level of 95% and a p < 0.05 was considered significant.
Among 419 patients, 266 (63.5%) patients received complete resection (CC-0) of macroscopic diseases. The overall survival curve is shown in Figure 1, and 5-and 10-year survival rates (-YSR) were 9.6% and 5.0%. Classification of Gastric Carcinoma [12]. The grade ranged from 0 to 3, with 0 indicating no evidence of effect; 1a, viable tumor cells occupy more than 2/3 of the tumor area; 1b, viable tumor cells occupy more than 1/3 but less than 2/3 of the tumor area; 2, viable tumor cells remain in less than 1/3 of the tumor area; grade 3, no viable tumor cells remain. In the study, grade 0 and 1a were considered histological non-responders, and grade 1b, 2, and 3 were considered as responders. All tumors were macroscopically classified according to the Bormann classification. After discharge, patients were followed up every 2 months until 3 years and every 6 months after 4-years as long as possible.
During follow up at the outpatient clinic, 184 patients were treated with postoperative systemic chemotherapy (40 regimens).

Ethical Standards
All subjects gave their informed consent for inclusion before they participated in the study. The study was conducted in accordance with the Declaration of Helsinki, and the protocol was approved by the Ethics Committee of Kishiwada Tokushukai and Kusatsu General Hospital ("A study of the safety and efficacy of a comprehensive treatment for the treatment of peritoneal metastasis from gastrointestinal cancer, H-19: Project identification code, on 26 October 2012"). All patients were informed about the adverse events of the procedure and gave their written informed consents to participate.

Evaluation of Complications
Postoperative complications were graded according to the system of classification reported by Dindo and colleagues [13].

Data Analysis
The survival analysis was performed using the Kaplan-Meier method and compared using the log rank test. Categorical variables were compared by X 2 analysis or Fischer's exact test. For multivariate analysis, a Cox regression model using factors which are significantly associated with death by univariate analyses was performed. Statistical analyses were performed using SPSS version 11.5 (SPSS Inc., Chicago, IL, USA). Confidence interval were calculated at the level of 95% and a p < 0.05 was considered significant.
Regarding the cytologic status before NIPS, patients with negative cytology survived significantly longer than those with positive cytology (p = 0.006, X 2 = 7.54). After NIPS, the survival difference was even more significant. Patients with negative cytology after NIPS survived significantly longer than those with positive cytology after NIPS (p < 0.0001, X 2 = 26.2).
Patients with positive cytology but no macroscopic PM (P0/Cy1), had a significantly better survival rate than patients with macroscopic PM (p = 0.0007, X 2 = 11.4).
Twenty-one patients survived longer than 5 years, and 15 patients are still alive without recurrence (Tables 4 and 5). All these patients received complete cytoreduction. Age of 21 5-year survivors ranged from 23 to 72 years old, and male and female were 13 and 8, respectively. Macroscopic types of primary tumors were 13 for type 4, 7 for type 3, and 1 for type 0 ( Table 4). Among these 21 patients, 7 have positive peritoneal cytology without macroscopic PM (P0Cy1) before NIPS, but post-NIPS PCI of these patients changed from 0 to 1 in 3 P0/Cy1 patients. Pre-and post-NIPS PCI ranged from 0 to 13, and 0 to 10, respectively. Pre-NIPS cytological status was positive in 13 patients, but only 2 patients showed positive peritoneal cytological status after. These 2 patients are still alive without recurrence 8 and 11 years after treatments.
Seven (29.2%) of 24 P0Cy1 patients survived longer than 5 years after treatment, and 2 patients died of recurrence 5 years after treatment. One patient received oophorectomy 4.5 years after the treatment due to ovarian recurrence and died of retroperitoneal recurrence after 10 years (Tables 4 and 5). Among 395 patients with macroscopic PM (P1), 14 (3.5%) patients survived longer than 5 years and 4 patients died of recurrence in one in the pancreas, 1 in the bone, 1 in the peritoneum, and 1 in the retroperitoneum.

Discussion
The present study clearly demonstrated that complete cytoreduction by peritonectomy procedures is considered to be essential for long-term survival [14][15][16]. MST, 5-YSR, and 10-YSR after complete cytoreduction were 20.5 months, 14.3%, and 8.3%, In contrast, all patients who received incomplete cytoreduction died within 7 years after treatment and MST, 5-YSR, and 10-YSR after incomplete cytoreduction were 12.0 months, 1.8%, and 0%, and the results were similar to those after systemic chemotherapy alone [4]. Accordingly, patients who are supposed to receive incomplete cytoreduction by exploratory laparoscopy and/or preoperative image diagnostic modalities, are not indicated for CRS, because of the high incidence of postoperative mortality and morbidity [8,9,11].
It is known that complete cytoreduction is one key element for increased oncologic outcome of patients with PM from gastric cancer. While interpreting our results, where we could not demonstrate a significant difference between patients treated with CC-0 compared to CC-1, we had to analyze the previous publications. Glehen et al. confirmed the completeness of cytoreduction score as a significant factor in multivariate analysis of their cohort [8]. In 2014, our group demonstrated the importance of completeness of cytoreduction as a significant factor in multivariate analysis comparing patients treated with CC-0 and CC-1 with CC-2 and CC-3 [11] The present study demonstrated that the PCI cutoff levels before NIPS are the more powerful prognosticator than completeness of cytoreduction. Prognosis of patients after cytoreduction could be strongly associated with the residual microscopic tumor burden [11,17]. These results may indicate that residual microscopic tumor burden in patients with PCI before NIPS higher than the cutoff level must be higher than those in patient with PCI lower than cutoff level even after complete cytoreduction. Therefore, it can be concluded that complete macroscopic cytoreduction is essential, but that patients with residual microscopic burden higher than the level that cannot be completely eradicated by intraoperative HIPEC and postoperative chemotherapy does not significantly lead to an improved overall survival even after CC-0 cytoreduction.
Regarding the prognostic factors except CCR score, the present study demonstrated that pre-PCI ≥14, post-PCI ≥12, post-SB-PCI ≥3, histologic non-responder status, lymph node involvement of pN2/pN3, positive cytology before and after NIPS, macroscopic PM, and number of involved peritoneal sectors ≥5 were significant factors indicating poor prognosis even after complete cytoreduction. Among these factors, multivariate analyses identified pre-PCI ≤13, and negative cytology after NIPS as powerful prognostic factors. MST and 10-year survival rate of patients with pre-PCI ≤13 were 22.8 months and 14.0%, and those with negative cytology after NIPS were 16.8 months, and 6.8%. These results suggest that the micrometastasis burden left in the residual peritoneum which could be completely eradicated by HIPEC and/or postoperative chemotherapy in patients with pre-PCI ≥14, post-PCI ≥12 or positive cytological, may be higher than that in patients with pre-PCI ≤13, post-PCI ≤12, or negative cytologic status.
Recently, Passot et al. reported the 10-year survival of GC patients with PM treated with CRS plus HIPEC [14]. In 127 patients, the 5-and 10 -YSR were 14% and 10% with two 10-year survivors without recurrence. They concluded that the combination of CRS and HIPEC may cure around 10% of GC-patients with PM. Our study results were almost the same as those of Passot et al. [14]. About 90% of patients who received comprehensive treatment died of recurrence in peritoneal cavity [15,16,[18][19][20]. Bando et al. reported that the 5-YSR of P0CY1 patients after D2 gastrectomy alone was 2% [17], and that the cause of recurrence must be due to invisible micrometastasis burden left after gastrectomy + D2 dissection. However, the prognosis of these patients was reported to be improved by gastrectomy with perioperative chemotherapy. Yamamoto et al. treated 10 P0Cy1 patients with neoadjuvant systemic chemotherapy (NAC) with S1 plus docetaxel and CDDP plus gastrectomy [21]. Cytology became negative in 8 of 10 patients treated with systemic NAC, and survival was better in patients with negative cytology after NAC than in patients with positive cytology. Survival improvement was reported in P0Cy1 patients treated with gastrectomy plus postoperative chemotherapy using S1 [21][22][23]. NIPS using S1 plus palitaxel or docetaxel showed a trend toward improvement in overall survival [24][25][26]. Additionally, simple washing with a high volume of saline (extensive intraperitoneal lavage: EIPL) may improve the survival of P0Cy1 patients [27].
These results indicate that a certain volume of micrometastases left after CRS in the peritoneal cavity could be completely eradicated by perioperative chemotherapy. Accordingly, micrometastasis in the peritoneal cavity should be reduced as much as possible before CRS using NIPS. Compared to systemic chemotherapy, NIPS is a more powerful killer of tumor cells in intraperitoneal micrometastasis [28]. The micrometastases located outside of the surgically resected area must be eradicated just after CRS by HIPEC and early postoperative intraperitoneal chemotherapy.
According to Bando, serum Carcinoembryonic antigen (CEA) levels were significantly higher in P0CY1 patients who had recurrence after curative gastrectomy than in those who did not have recurrence. These results strongly suggest that patients with recurrence after CRS must have larger residual tumor burden which could be completely eradicated by intraoperative and postoperative chemotherapy.
To date, no study has reported the long-term survival of patients after comprehensive treatment with special reference to PM involving peritoneal sectors and abdominal organs. The purpose of the present study was to clarify the effect of removing infiltrating peritoneal sectors and organs on postoperative survival. Long-term survival could be obtained after surgical removal of PM-involved areas including 13 peritoneal sectors and abdominal organs except pancreas capsule, or ureter. The 10-YSR of patients with involvement of paracolic gutter, spermatic cord or sac, rectum, uterus, ovary, upper jejunum, and lower jejunum were 13.9%, 21.9%, 9.5%, 6.3%, 6.7%, 5.5%, and 5.6%, respectively. Patients with metastasis of diaphragmatic peritoneum and ileum who were disease-free for more than 5-years did not survive 10 years after treatment.
If the PM is found on the peritoneal surface or organs except pancreas and ureter, and the pre-PCI ≤ 14, post PCI ≤ 11, post SB-PCI ≤ 2 and number of involved peritoneal sector ≤5 after NIPS are confirmed, CCR-0 resection should be performed. Some patients with non-response after NIPS, and positive cytology after NIPS survived longer than 10 years but their survival rates were significantly lower than those with response to NIPS, and negative cytology after NIPS.
One of the limitations of this study is that it only included patients treated by one surgical group in two institutions, where patients were selected, and oncologic treatment plans recommended on the basis of our long experience in peritoneal surface malignancies. In this setting, selection bias was inevitably a factor in patient enrollment. The most important factors in patient selection were PCI, tumor response and biology, and performance status. Furthermore, due to the nature of a retrospective analysis, the treatment plan for some patients differed, including factors such as pretreatment PCI, the usage of HIPEC, and postoperative chemotherapy.

Conclusions
From the present data, PM should be removed aggressively to achieve complete cytoreduction for PM, when the patients have PCI after NIPS ≤ 11, PCI before NIPS ≤ 13, SB-PCI ≤ 2, involvement of ≤5 peritoneal sectors.
Only one randomized controlled study confirmed the effects of CRS plus POC [29]. New methods to more completely eradicate the residual tumor burden after CRS should be developed in the future.