Survival outcomes after cytoreductive surgery and hyperthermic intraperitoneal chemotherapy for peritoneal carcinomatosis from gastric cancer: a systematic review

Abstract

Background: The current treatment of choice for peritoneal carcinomatosis from gastric cancer is systemic chemotherapy. Cytoreductive surgery (CRS) and hyperthermic intraperitoneal chemotherapy (HIPEC) is a new aggressive form of loco-regional treatment that is currently being used in pseudomyxoma peritoneii, peritoneal mesothelioma and peritoneal carcinomatosis from colorectal cancer. It is still under investigation for its use in gastric cancer.

Methods: The literature between 1970 and 2016 was surveyed systematically through a review of published studies on the treatment outcomes of CRS and HIPEC for peritoneal carcinomatosis from gastric cancer.

Results: Seventeen studies were included in this review. The median survival for all patients ranged from 6.6 to 15.8 months. The 5-years overall survival ranged from 6 to 31%. For patients with complete cytoreduction, the median survival was 11.2 to 43.4 months and the 5-years overall survival was 13 % to 23%. Important prognostic factors were found to be a low peritoneal carcarcinomatosis index (PCI) score and the completeness of cytoreduction.

Conclusion: The current evidence suggests that CRS and HIPEC has a role to play in the treatment of peritoneal carcinomatosis from gastric cancer. Long term survival has been shown for a select group of patients. However, further studies are needed to validate these results.

Introduction

Despite a decreasing incidence worldwide, gastric cancer is the fifth most common cancer and the third leading cause of cancer-related death [1]. Peritoneal carcinomatosis (PC) from gastric cancer is an increasingly common finding in both the synchronous and metachronous setting. Patients with peritoneal metastasis have a worse prognosis than patients with metastatic disease at other sites [2]. Without treatment, the median survival in patients with peritoneal carcinomatosis from gastric cancer is 1–3 months [3, 4].

Palliative chemotherapy is beneficial when compared to best supportive care. Average median overall survival ranges from 4.3 months to 11 months, and is better with a combination of chemotherapy agents [5]. Even more, salvage chemotherapy leads to a significant reduction in the risk of death when compared to best supportive care [6, 7]. Data analysis from palliative chemotherapy results show an overall survival of 8–12 months [8–13]. These rates are not significantly improved despite the addition of newer agents (docetaxel, S1) and targeted therapy [14–18]. With PC being a common finding and the results for systemic chemotherapy being unsatisfactory, this has led many to search for new treatment options for these patients. The basis for treating these patients with peritoneal carcinomatosis so aggressively is that peritoneal disease should be considered locoregional spread, rather than metastatic disease.

Principles of cytoreductive surgery (CRS) and hyperthermic intraperitoneal chemotherapy (HIPEC)

CRS and HIPEC are being increasingly used worldwide for patients with peritoneal carcinomatosis. It is the standard of care for patients with PMP and mesothelioma and is becoming widely accepted for patients with peritoneal carcinomatosis from colorectal cancer [19]. As patients with PC from gastric cancer have limited treatment options with median overall survival less than 12 months, CRS and HIPEC has been increasingly used in this field. The first documented report of CRS and HIPEC for peritoneal carcinomatosis in gastric cancer was in 1988 by Fujimoto [20]. They operated on 15 patients with advanced gastric cancer and peritoneal disease and they concluded that CRS combined to HIPEC was safe and well tolerated. Since then, many units have gone on to investigate the use of CRS and HIPEC for PC from gastric cancer. As the diagnosis of PC increases and the use of CRS and HIPEC increases, this review is important to determine the effectiveness of CRS and HIPEC in PC from gastric cancer as defined by the long term survival.

Materials and methods

Search strategy

A systematic review of the literature was conducted by two independent authors (CSC, RAS) using PubMed, Medline, Embase, Cochrane database and Ovid search. Combinations of the following search terms were used to identify the studies: “gastric cancer”, “gastric adenocarcinoma”, “peritoneal carcinomatosis”, “peritoneal disease”, “cytoreductive surgery”, “peritonectomy”, “hyperthermic intraperitoneal chemotherapy”, “intraperitoneal chemotherapy”. Articles were first selected based on their titles. Further elimination was performed based on the abstracts. These last articles went through a full-text selection.

Study inclusion and data extraction

All retrospective or prospective studies investigating the use of CRS and HIPEC for peritoneal carcinomatosis from gastric cancer since 1970 were eligible for inclusion in this systematic review. Case reports were excluded. Articles were only included if they reported survival outcomes. Only articles that had at least 10 patients were included. In presence of heterogeneous group of tumour types, only articles reporting separate outcomes for gastric cancer were included. Articles that studied CRS and HIPEC in the prophylactic setting but reported separate outcomes for those with peritoneal disease were included. The search process is summarized in Figure 1.

Figure 1:

Prisma flowchart (articles included for survival outcomes).

Data extraction

An initial evaluation of the title, abstract and keywords of every record found was performed by CSC and RAS. The next step was the retrieval of the full text of potentially relevant trials. The two reviewers independently assessed the eligibility of each potentially relevant trial with the use of inclusion criteria.

The information extracted included the following: (1) General characteristics of the studies (first author and year of publication of the article, country, number and age of the participants, inclusion and exclusion criteria; (2) Design and characteristics of the study (3) Main results: Overall and disease free survival for all patients as well as the subset of patients with complete cytoreduction; (4) Morbidity and mortality; (5) Prognostic factors including extent of the disease; (6) HIPEC techniques.

Results

Study selection

Literature search using the search strategy described above identified 224 studies. After selecting for humans only studies, those published in English and those published after 1970, only 145 studies remained. The abstracts of these 145 articles were reviewed carefully and a further 42 articles had full text review. Ultimately, 17 articles were included in this review and summarized in Table 1 [20–37]. Of the 17 articles included in the review, 1 was a systematic review, 1 was a randomized controlled trial, 11 were prospective studies (7 case series, 4 case controls and 1 prospective randomized trial) and the remaining 4 were retrospective studies.

Table 1:

Survival outcomes after CRS and HIPEC for peritoneal carcinomatosis from gastric cancer.

Author	Year	Type of study	No. of patients	Control arm	Outcome	Outcome (CC0/1)
Yonemura	1991	Prospective	41	N.A.	Median overall survival 14.6 months (range 1–64.2) 3 years survival: 28.5%	NR
Yonemura	1996	Prospective	83	N.A.	5 years survival 11%
Fujimoto	1997	Prospective	48	18 (surgery only)	HIPEC vs control (p=0.001) 5 years survival 31% vs 0% 8 years survival 25% vs 0% Death due to peritoneal recurrence (HIPEC vs control) 27% vs 94%	NR
Fujimoto	1999	Prospective randomised	71	70 (surgery only)	HIPEC vs control (p=0.04) 4 years survival 76% vs 58% 8 years survival 62% vs 49%	NR
Hirose	1999	Prospective case control	17	20 (surgery only)	HIPEC vs control (p=0.04) 1 years survival 44.4% vs 15.8% Median survival 11 months vs 6 months	NR
Glehen	2004	Prospective	49	N.A.	Median overall survival 10.3 months 5 years survival 16%	Overall median survival 21.3 months
Hall	2004	Propsective case control	34	40 (no peritoneal disease)	HIPEC group Median overall survival 8 months 5 years survival 6%	R0/1 vs R2 Median overall survival 11.2 months vs 3.3 months 2 years survival 45% vs 8%
Yonemura	2005	Retrospective	42	65 (conventional surgery + HIPEC)	All patients Median survival 11.5 months 5 years survival 6.7	Complete vs incomplete 5 years survival 13% vs 2% Median survival 19.2 months vs 7.8 months
Scaringi	2008	Retrospective	26	N.A.	Median survival 6.6 months	CC0 vs CC2 (p=0.007) Median survival 15 months vs 3.9 months
Glehen	2010	Retrospective	159	N.A.	Median overall survival 9.2 months 5 years survival 13%	Median overall survival 15 months 5 years survival 23%
Yang	2010	Prospective	28	N.A.	2 years survival rate 42.8%	CC0 vs CC1 vs CC2/3 Median survival 43.4 months vs 9.4 months vs 8.3 months
Yang	2011	RCT	34	34 (surgery only)	HIPEC vs controlMedian survival 11 months vs 6.5 months 3 years survival 5.9% vs 0%	CC0/1 vs CC2/3 Median survival 12 months vs 8.2 months
Gill	2011	Systematic review	441	N.A.	Median overall survival 7.9 months 5 years survival 13%	Median overal survival 15 months
Canbay	2013	Prospective	152	N.A.	Median survival 15.8 months 5 years survival 10.7%	CC0/1 better survival HR 1.55 95% CI 1.04–2.2 p=0.03
Magge	2014	Prospective	23	N.A.	Median overall survival 9.5 months 3 years overall survival 18% 3 years disease free survival 10.7%	Complete cytoreduction significant on univariate and multivariate analysis for overall survival
Chia	2016	Retrospective	81 (all CC0/1)	N.A.	N.A.	Median overall survival 17.3 months 5 years overall survival 18% 5 years disease free survival 11%

NR, not reported.

Outcomes after CRS and HIPEC

Overall survival

All seventeen studies reported the overall survival. Reported outcomes include median overall survival, 1-, 2-, 3-, 4-, 5-, 8-years survival. Overall median survival ranged from 6.6 to 15.8 months. 1-year survival rate was 44.4%. 2-years survival rate was 43 % to 45% 3-years survival rate ranged from 5.9% to 28.5% 4-years survival rate was 76%. 5-years survival rate ranged from 6% to 31%. In the 3 case control studies that used surgery only as the comparative group, overall survival in the HIPEC group was better than that in the control group [24–26]. This was statistically significant.

Eleven studies reported overall survival in patients with complete cytoreduction separately. Overall median survival for this group of patients ranged from 11.2 to 43.4 months. One paper reported 2 years overall survival of 45% while 2 reported a 5-years survival of 13 % to 18%.

Disease-free survival

Only two of the 17 studies reported the disease free survival. Magge et al. reported a 3-years disease free survival of 10.7% [36]. Chia et al. reported a 5 years disease free survival of 11% but in the subset of patients with complete cytoreduction [37].

Morbidity and mortality

Fourteen of the 17 studies reported their morbidity rates and are summarized in Table 2. The morbidity rate ranged from 2.8% to 52.2%. The systematic review by Gill et al. reported a morbidity rate of 21.5%. One complication of concern is the rate of anastomotic leak after gastric resection. Several papers have addressed this concern in CRS and HIPEC for gastric cancer as well as other pathologies [31, 38–41]. In these papers, there were no reported cases of leak after gastric resections.

Table 2:

Mortality and morbidity rates after CRS and HIPEC.

Author	Year	No. of patients	Mortality rate, n (%)	Morbidity rate, n (%)
Yonemura	1991	41	0	12
Fujimoto	1999	71	NR	2.8
Hirose	1999	17	5.8	35.2
Glehen	2004	49	4	27 (major)
Hall	2004	34	0	35
Yonemura	2005	42	7	43
Scaringi	2008	37	5.4	27
Glehen	2010	159	6.5	27.8 (major)
Yang	2010	28	0	14.3
Yang	2011	34	NR	14.7
Gill (systemic review)	2011	441	4.8	21.5
Canbay	2013	152	3.9	23.6
Magge	2014	23	4.3	52.2 (major, 60 day)
Chia	2016	81	2.5	44 (major)

The mortality rate was reported in 12 papers and this ranged from 0% to 7%. In a systematic review by Gill et al., the mortality was found to be 4.8% [34].

Prognostic factors

Eleven out of seventeen studies did an analysis of the factors that may affect survival. The 2 most commonly found factors were the extent of disease and the completeness of cytoreduction.

The extent of disease was found to be an important prognostic factor in 5 studies [30–32, 35, 37]. Scaringi et al. showed a survival difference between those with Gilly stage 1 and 2 vs those with stage 3 and 4 (15 vs 4 months, p=0.014) [30]. Glehen et al. showed in 2010 that in patients with complete cytoreduction, the PCI was the only independent prognostic factor [31]. If the PCI was > 19, no patients survived more than 6 months and if the PCI was > 12, no patients survived more than 3 years. Yang et al. showed a survival difference between those with a PCI≤20 and those with PCI≥20 [32]. Those with PCI≤20 had a median overall survival of 27.7 months vs 6.4 months in those with PCI > 20 (p=0.0001). Canbay et al. showed that a PCI≤6 was an independent prognostic factor for their patients undergoing CRS and HIPEC after bi-directional chemotherapy (HR 2.16 95% CI, p=0.013) [35].

This was also seen in the Chia et al. study found that a PCI of < 7 was a significant predictor of survival [37]. Those with PCI < 7 had a median overall survival of 26.4 months compared to 10.9 months of those who had a PCI≥7 (HR 2.67, 95% CI, 1.54–4.64, p < 0.001). All the patients who were disease free at 5 years had a PCI < 7. This same PCI cut-off was seen in a study by Yonemura et al., who found a PCI < 7 predicted for improved survival (median survival 2.8 years vs 1.1 years p=0.0001) [42].

The completeness of cytoreduction was found to be a prognostic factor in 11 studies [26–33, 35–37]. Glehen et al. showed in 2004 that with a CC score of 0 or 1, the median overall survival improved from 10.3 months to 21.3 months [27]. The 5-years overall survival improved from 16% to 29.4%. The multi-center study in 2010 showed an improvement of median overall survival of 9.2 months to 15 months if CC0 was achieved [31]. Five years overall survival also improved from 13% to 26%. On multivariate analysis, the completeness of cytoreduction was the only independent variable (RR=2.04, p < 0.001). Yonemura et al. showed a 2.08 increased risk of dying if complete cytoreduction was not performed [42].

Other prognostic factors that were reported include the presence of ascites, synchronous disease, chemotherapy of more than 6 cycles, the absence of complications, lymph node involvement and the presence of 2 or more anastomosis [27, 33, 35, 36].

The signet ring cell (SRC) histology was reported in six studies and analyzed as possible prognostic factors in two studies. Magge et al. found that the presence of SRC was not a predictor for overall survival but a predictor for disease-free survival even on multivariate analysis (p=0.001) [36]. Chia et al. found a trend towards better overall survival for patients without SRC but this was not statistically significant [37].

HIPEC techniques

One issue that one has to take into account when interpreting the results of these trials for peritoneal carcinomatosis in gastric cancer is that these trials are very heterogeneous. There are variations, especially in terms of the conduct of the HIPEC, in all these trials. These range from the method of HIPEC, the chemotherapy used, the dose of the drug, the duration of the HIPEC to the temperature used. These variations are shown in Table 3. As each institution has its own method of HIPEC, the answer as to which method is superior may not be answered until we have further studies that look into this.

Table 3:

Variations in HIPEC.

Author	Year	Method	Drug	Dosage	Duration (mins)	Temperature (°C)
Yonemura	1991	Closed	Mitomycin	5 μg/mL	40–60	41–43
			Cisplatin	30 μg/mL
Yonemura	1996	Closed	Mitomycin	30 mg	60	42–43
			Cisplatin	300 mg
			Etoposide	150 mg
Fujimoto	1997	Closed	Mitomycin	10 μg/mL	120	43–45
Fujimoto	1999	Closed	Mitomycin	10 μg/mL	120	43–45
Hirose	1999	Open	Mitomycin	20 mg	50	41–44.5
			Cisplatin	100 mg
			Etoposide	100 mg
Glehen	2004	Closed	Mitomycin	40–60 mg	90	46–48
Hall	2004	Closed	Mitomycin	40 mg	120	40
Yonemura	2005	Closed	Mitomycin	30 mg	60	40–42
			Cisplatin	300 mg
			Etoposide	150 mg
Scaringi	2008	Open	Mitomycin	120 mg	90–120	41–43
			Cisplatin	200 mg/m2
Glehen	2010	Open 46.5%	Mitomycin	30–50 mg	60–120	41–42.5
		Closed 53.5%	± Cisplatin	50–100 mg/m2
			Oxaliplatin	360–460 mg/m2	30	43
			± Irinotecan	100–200 mg/m2
			± 5FU and leucovorin
Yang	2010	Open	Hydroxycamptothecin	20 mg	90–120	42.5–43.5
			Or cisplatin	120 mg
			+ Mitomycin	30 mg
Yang	2011	Open	Cisplatin	120 mg	60–90	42.5–43.5
			Mitomycin	30 mg
Canbay	2013	Open	Docetaxel	30 mg/m2
Magge	2014	Closed	Mitomycin	40 mg	100	42
Chia	2016	Open and closed	Mitomycin 87.5%	NR	NR	NR
			Cisplatin 10%
			Oxaliplatin 2.5% Clo

Discussion

At the initial diagnosis, synchronous peritoneal carcinomatosis can occur in up to 43% of gastric cancer cases [3]. PC accounts for 35% of synchronous metastasis and is the sole site of metastasis in 9% of cases [3]. In the recurrent setting, peritoneal carcinomatosis can occur in up to 46% of cases and as the only site in up to 40% of cases [3, 43–48].

The current treatment option is palliative chemotherapy. Many of them are treated with palliative chemotherapy which has been shown to give an overall survival of 8 to 12 months [8–13]. These studies were performed in patients with metastatic disease at all sites. The theory is that tumours in the peritoneum do not respond as well to the chemotherapy. This is partly because of the plasma peritoneal barrier which reduces the penetration of the systemic chemotherapy to the tumours in the peritoneum [49]. In addition, the poor vascularity of the peritoneum and the poor oxygenation of these tumours result in a low apoptotic potential of these cancer cells [49, 15]. Hence, the response rates of patients with peritoneal only disease is about 14 % to 25% [13, 50, 51]. The median survival of patients with PC undergoing palliative chemotherapy is 9.5 to 12 months [52, 53].

The SRC histology portends a worse prognosis. Piessen et al. showed that the median survival was significantly lower for SRC patients (21 vs 44 months, p < 0.004), with higher rates of localized peritoneal carcinomatosis (p < 0.013) and lymph node involvement (p < 0.001) at diagnosis, lower R0 resection rate (p < 0.019) and earlier tumor relapse (p < 0.009), which was generally in a peritoneal carcinomatosis form (p < 0.011) [54]. Furthermore the SRC histology proved to be chemoresistant [55].

Therapeutic strategy in SRC gastric cancer is difficult to evaluate due to the scarcity of data specifically analyzed. One retrospective study done on 18 patients with SRC gastric cancer by Königsrainer et al. showed a median survival of 8.9 months within a median follow-up of 6.6 (0.5~31) months [56]. Of note, complete cytoreduction could only be achieved in 72% of these patients with a progression-free survival of 6.2 months in this subgroup [56].

These 18 patients had preoperative chemotherapy and a high rate of non resectable patients, probably due to the absence of preoperative staging laparoscopy [56].

This subgroup of patients should be independently analyzed.

In 1980, Spratt was the first to report the use of CRS and HIPEC in a case of pseudomyxoma peritoneii [57]. This technique was further popularized by Dr Paul Sugarbaker [58]. The main principle of surgery is the complete removal of all macroscopic or visible disease. Immediately after this step, the microscopic disease is then treated with HIPEC. During surgery, it is imperative that all tumour greater than 1 mm is removed as it has been shown that the chemotherapy cannot penetrate beyond 1–2 mm in depth [59, 60]. The effect of adding hyperthermia has been shown in vitro and in vivo to potentiate the penetration and activity of the chemotherapeutic agent [61, 62]. This step is done immediately after surgery so that the chemotherapy can access all the planes in the abdomen. Subsequent to surgery, adhesions form and the tumour cells may become entrapped, thereby reducing the possibility of the chemotherapy reaching them [63, 64].

The rationale for HIPEC is multi fold. The previously mentioned plasma peritoneal barrier has a role to play in allowing us to administer higher doses of chemotherapy intraperitoneally than we would intravenously. It maintains a high intraperitoneal concentration and a low plasma concentration and creates a positive gradient [65]. The hyperthermia enhances the intraperitoneal chemotherapy directly and indirectly. The direct cytotoxic effect impairs DNA repair, denatures proteins, inhibits oxidative mechanisms and increases lysosomal activities in the tumour cells [65–67]. Indirectly, it increases the penetration and uptake of the drug into the tumour cells as well as the chemosensibility of the cells [65, 67, 68].

Our review has shown there is a role for CRS and HIPEC in patients with PC from gastric cancer. There is improved survival over patients undergoing surgery alone. However, there is a select group of patients that benefit the greatest from this treatment modality. From this review, it is clear that the most important factor that improves survival is the completeness of cytoreduction. The extent of disease is also an important prognostic factor although the exact cut-off PCI score is yet to be determined.

One concern regarding patients undergoing CRS and HIPEC is the quality of life that these patients have after the surgery. The surgery is presumed to have a lasting negative impact on the patients quality of life. While there have been no studies that look at patients undergoing CRS and HIPEC for gastric cancer in particular, there have been numerous retrospective and prospective studies that have looked at patients undergoing CRS and HIPEC for various pathologies [69–76]. Various quality of life instruments have been used which include the EORTC and SF 21. All these papers show a similar trend, which is that the QOL drops from baseline immediately after surgery but recovers to baseline after 6–12 months. Not only the global health, physical scores and symptoms improve but large increases were also seen in the emotional health of patients [77, 78]. There is a definite psychological component as these patients were likely previous told that they were palliative and after undergoing CRS and HIPEC and being given a chance at long term survival contributes to their emotional wellbeing.

Patients that are at risk of developing peritoneal metastasis after curative resection for gastric cancer include those with serosal invasion, advanced nodal stage and those with positive cytology [46, 79, 80]. Positive cytology constitutes M1 disease in the latest AJCC 7th edition staging [81]. Patients who have positive cytology have a 81% chance of peritoneal metastasis after curative surgery as opposed to 45% of those who have a negative cytology [42]. Chia et al. also showed that patients who present with their peritoneal disease synchronously and had CRS and HIPEC did better than those who present metachronously [37]. Hence the idea of giving HIPEC at the initial curative surgery is gaining traction.

This was first reported in 1988 by Koga et al. [82]. They looked at gastric cancer patients with serosal invasion; 38 of them underwent surgery and HIPEC while 55 only had gastric surgery. The HIPEC group had a better 3-years overall survival of 74% to 53% in the group that did not have HIEPC (p=0.04). The HIPEC group also had less peritoneal recurrences (36%) compared to the group that did not have HIPEC (50%).

Since then, various Asian groups have gone on to investigate the use of HIPEC in the prophylactic setting [21, 25, 26, 83–87]. All of them have provided evidence to show that HIPEC in the prophylactic setting can reduce the incidence of peritoneal recurrence and improve survival. However, HIPEC in the prophylactic setting for gastric cancer is still not standard practice. There is a lack of data from the western countries where the incidence of gastric cancer is lower. There are currently 2 ongoing trials: GASTRICHIP which is being run in France and another trial run by the European Network of Excellence (EUNE) [88, 89]. Both trials are randomizing patients at high risk of developing peritoneal disease after curative resection for gastric cancer into HIPEC or no HIPEC after the initial curative surgery.

Another concept that is being investigated in the treatment of PC from gastric cancer is bi-directional chemotherapy. Bi-directional chemotherapy works by simultaneously delivering systemic and intraperitoneal chemotherapy, thereby creating a diffusion gradient and creating a wider treatment area. The aim of this is stage reduction, the removal of free intraperitoneal cancer cells and to increase the chance of complete cytoreduction. The chemotherapy is delivered via a combination of oral, intravenous and intraperitoneal routes. There has been recent interest in bi-directional chemotherapy from worldwide groups [33, 35, 39, 90–93].

In a Phase I/II study of intraperitoneal docetaxel and S1 by Fushida et al., they showed that 52% of patients had peritoneal disease that responded to the chemotherapy [93]. 11% was complete response while 41% was partial response. 37% had stable disease. Of the 22 patients that had a positive peritoneal cytology, 18(82%) of them turned negative. In these trials, only patients who had a response to bi-directional chemotherapy went on to have surgery. This ranged from 53% to 85% [35, 42, 92–94]. 65–70.7% of patients had complete cytoreduction. Bi-directional chemotherapy shows promise as tools to help us better select the patients that may be suitable for CRS and HIPEC.

Conclusion

Numerous papers have been published on CRS and HIPEC for PC from gastric cancer. Good and improved long-term survival has been shown in a select group of patients. In light of published literature, we suggest that CRS and HIPEC can be considered in patients with no evidence of distant metastasis, a low volume of peritoneal disease and in whom complete cytoreduction can be achieved (CC0 or 1). Further studies need to be done to investigate the variability of the HIPEC, the use of HIPEC as a prophylactic tool and the use of bi-directional chemotherapy to select patients for CRS and HIPEC.