Emerging treatment modalities for gastric cancer with peritoneal metastases: a systematic review

Abstract

Introduction

Gastric cancer remains associated with a high mortality, since patients often present with advanced-stage disease such as peritoneal metastases. Treatment options in this setting are limited and mainly consist of palliative systemic therapy.

Objectives

This systematic review aims to summarize emerging therapeutic strategies for patients with peritoneally metastasized gastric cancer in relation to response, toxicity, and survival.

Methods

A search was performed in PubMed, EMBASE, and Cochrane database until December 8, 2023. Studies were eligible if they prospectively investigated any form of cancer treatment in adult patients with peritoneally metastasized gastric cancer.

Results

In total, 25 articles were retrieved from 9 different countries, investigating systemic therapy, intraperitoneal therapy (IP), hyperthermic intraperitoneal chemotherapy (HIPEC) with or without cytoreductive surgery (CRS), and pressurized intraperitoneal aerosol chemotherapy (PIPAC). CRS-HIPEC followed by early postoperative intraperitoneal chemotherapy (EPIC) showed high complication rates and patient drop-out. Median overall survival (OS) ranged from 6.5 to 7.4 months with systemic chemotherapy, 13.0 to 23.9 months with IP chemotherapy, 6.8 to 13.0 months after PIPAC, 11.0 to 19.0 months with CRS-HIPEC, and 9.9 months in elderly with HIPEC alone. Notably, patients with a Peritoneal Cancer Index ≥20 showed median OS ≥13.0 months with IP chemotherapy or HIPEC without CRS.

Conclusion

Few prospective trials are conducted in patients with peritoneally metastasized gastric cancer. Although IP chemotherapy and CRS-HIPEC show encouraging results in selected cohorts, especially in Asia, current evidence from randomized studies is limited. Further prospective trials are needed to clarify their role alongside the current systemic therapy, including immunotherapy and targeted therapy.

Implications for Practice.

To date, few prospective clinical trials specifically focus on patients with peritoneally metastasized gastric cancer. While intraperitoneal chemotherapy and CRS-HIPEC have shown encouraging results in selected cohorts, many trials fail to report the Peritoneal Cancer Index (PCI) scores or lack consistent use of and comparison to modern systemic therapy regimen, compromising generalizability of the outcomes. This review highlights the need for well-designed prospective studies on intraperitoneal chemotherapy, CRS-HIPEC, and systemic therapy that incorporate PCI-based stratification and compare outcomes against up-to-date systemic treatments, including immunotherapy and targeted therapy, to develop reliable treatment guidelines for patients with peritoneally metastasized gastric cancer.

Introduction

Although the overall survival (OS) rate of patients with early-stage gastric cancer has improved over the last decades, the mortality rate of patients with metastatic gastric cancer remains high.^1-3 The 5-year survival rates range from 70% in patients with early-stage gastric cancer to 6% in patients with metastatic gastric cancer.¹ At the time of diagnosis, approximately 40% of patients present with metastatic disease, with the peritoneum as the most frequent metastatic site.^2,3 The survival is negatively related to the extent of peritoneal dissemination.² Dutch population-based data show a median OS of 7.3 months (range 4.0-12.5) for synchronous peritoneal metastases with any type of tumor-directed therapy.⁴

Treatment options are currently limited in the case of metastatic disease and consist merely of palliative systemic therapy. However, systemic chemotherapy has been shown to be less effective in patients with peritoneally metastasized gastric cancer, possibly due to the plasma-peritoneal barrier that prevents chemotherapy from entering the peritoneal cavity.^2,3,5 New treatment modalities have been developed aiming to elevate drug concentrations in the intraperitoneal compartment. Potentially new treatment options include intraperitoneal (IP) chemotherapy, hyperthermic intraperitoneal chemotherapy (HIPEC) (with/without cytoreductive surgery [CRS] and/or early postoperative intraperitoneal chemotherapy [EPIC]), pressurized intraperitoneal aerosol chemotherapy (PIPAC) and targeted therapy. Several studies showed that these treatment modalities are safe and effective in other peritoneally metastasized malignancies.^6–9 Currently, the effect of these new treatment modalities is studied in strictly selected populations and may change the perspectives of patients with peritoneally metastasized gastric cancer.

This systematic review assesses the effect of current and new therapeutic options for patients with peritoneally metastasized gastric cancer on survival and provides an estimation on what treatment modality can benefit patients with either limited or extensive PM.

Materials and methods

This systematic review was registered in the international prospective register of systematic reviews (PROSPERO) and conducted according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines. The research was performed independently by 2 researchers (M.T.W. and L.G.), who both executed the literature search, study selection, data extraction, risk of bias assessment, and data synthesis. In case of disagreement, a third investigator (Iv.H.) made the definite statement.

Study selection

Prospective clinical trials treating adult patients with peritoneally metastasized gastric cancer were selected, such as prospective cohort studies, feasibility trials, single-arm interventional trails, and randomized controlled trials (RCTs). Studies were included if they presented results on any type of anti-cancer treatment, for example, intravenous chemotherapy, radiotherapy, primary tumor resection, immunotherapy, IP chemotherapy, CRS-HIPEC, PIPAC, or EPIC. Studies were only included if a full-text English version was available and if published in the years 2000 to 2023. In case of overlapping publication, only the most complete articles were included. For subgroup analysis, we selected the trials presenting the peritoneal cancer index (PCI). Studies were excluded if they solely described emergency treatment, such as emergency resection or emergency radiotherapy (eg, in case of perforation or severe bleeding). Duplicate publications were excluded.

Search strategy

A search was performed using the PubMed, EMBASE, and Cochrane databases until December 2023. The databases were systematically searched with a date restriction from January 1, 2000 to December 8, 2023. For future trials, ClinicalTrial database was searched until December 8, 2023. The full search queries are presented in Supplementary Appendix. The references of all eligible manuscripts were searched for additional eligible studies.

Selection process

After de-duplication, all articles obtained from the search were screened on title and abstract by 2 reviewers (M.T.W. and L.G.), using Rayyan database (Rayyan Systems Inc., Cambridge, MA, USA). Articles were excluded if the studied population or intervention did not evidently correspond with the domain of this review. All articles that adhered to the eligibility criteria were included.

Data collection

We extracted the following data: date of publication, author, country, enrollment period, number of participating centers, inclusion criteria, total number of patients included, study design, treatment modality, and, if applicable, dosage and treatment schedule, concomitant treatment, extend of the PM, median OS, safety, and details relevant to the risk of bias. If available, additional data were extracted on the extend of the PM.

Risk of bias

Both reviewers assessed the risk of bias by using the Cochrane risk-of-bias tool for all included RCTs, the ROBINS-I Risk Of Bias In Non-randomized Studies of Intervention in all phase I and II trials, and the Newcastle-Ottawa scale in case–control and cohort studies.

Results

The search identified 531 records from the databases, containing 416 unique articles after removal of duplicates. After screening the title and abstract, 108 articles were retained for full-text screening by 2 independent reviewers (M.W. and L.G.). In the process of full-text screening, we included another 3 articles by reference screening. Articles were excluded for the following reasons: a different study population, a wrong study design, absence of a full-text English version, or different outcome measures. Finally, a total of 25 trials were included, investigating treatment strategies in 1291 patients with peritoneal gastric cancer and PM (Figure 1). Treatment modalities studied in these trials are systemic therapy, IP chemotherapy, HIPEC (either with or without CRS), and PIPAC. Therefore, we described these different modalities separately in the different sections of this review.

Figure 1.

Flow chart of study selection.

Systemic therapy

Two trials were included that specifically investigated the effect of systemic therapy in 493 patients with peritoneally metastasized gastric cancer (Table 1).^10,11 Shitara et al. (2017)¹⁰ randomly assigned patients with metastatic gastric cancer to receive either three-weekly nab-paclitaxel, weekly nab-­paclitaxel, or solvent-based paclitaxel (1:1:1), with stratification for the presence of PM in the randomization process. In addition, subgroup analysis was performed for patients with PM. Nakajima et al. (2020)¹¹ investigated the effect of weekly 5-FU (600 mg/m²) and l-leucovorin 5-FU (250 mg/m²) compared with weekly 5-FU (500 mg/m²), l-leucovorin (250 mg/m²), and paclitaxel (60 mg/m²) (FLTAX) in patients with gastric cancer and severe PM, expressed as massive ascites and/or inadequate oral intake. In both trials, no PCI was reported.

Table 1.

Overview of trials conducting systemic therapy.

Author (year)	Country	Design	GCPM n =	Inclusion criteria	Median age (range)	Median PCI (range)	Trial treatment	AEs (grade ≥ 3)	Median OS in months	Clinical response rate
Shitara (2017)  10	Japan	Phase III RCT	131 vs 131 vs 130	Stage IV GC refractory to a fluoropyrimidine-containing first-line CTx regimen.	md	md	3-weekly nab-PTX vs weekly nab-PTX vs weekly solvent-based PTX	md	md	Md
Nakajima (2020)  11	Japan	Phase II/III RCT	51 vs 50	GCPM with unresectable or recurrent disease. Specifically for patients with either severe PM, massive ascites and/or inadequate oral intake.	64 (41-75) vs 65 (29-75)	md	Weekly 5-FU-/l-LV vs weekly FLTAX	78.4% vs 77.1%	6.1 vs 7.3 (P = .14)	85% vs 70% (P = .041)

Abbreviations: 5-FU = 5-Fluorouracil; AEs = adverse events; CTx = chemotherapy; FLTAX = 5-FU, Leucovorin, Paclitaxel; GC = gastric cancer; GCPM = gastric cancer patients with peritoneal metastases; l-LV = leucovorin; md = missing data; nab-PTX = nanoparticle albumin-bound paclitaxel; OS = overall survival; PCI = Peritoneal Cancer Index; PM = peritoneal metastases; RCT = randomized controlled trial.

Median OS was 6.1 months (95% CI 3.4-7.8) with FL vs 7.3 months (95% CI 5.0-10.6) with FLTAX (P = .14). The median PFS was 1.9 months (95% CI 1.5-3.5) with FL and 5.4 months (95% CI 2.6-6.9) with FLTAX (HR 0.64; P = .029). No median OS was reported specifically for peritoneally metastasized gastric cancer in the trial of Shitara (2017), although subgroup analysis in this population showed a better median OS and PFS in patients receiving weekly nab-paclitaxel compared to those receiving solvent-based paclitaxel.

Grade III-IV adverse events were observed in 78.4% and 77.1% of patients in the FL- and FLTAX-arm, respectively. Most commonly observed adverse events were anorexia (47% vs 31%), leukopenia (28% vs 13%), and neutropenia (30% vs 21%), all more frequently observed in the FL-arm. In addition, 2 treatment-related deaths (3.9%) occurred in the FL-arm, within 30 days of randomization.¹⁰ Quality of life (QoL), oral intake, and drainage free survival seemed to be favorable in FLTAX compared to FL, although this was not proven statistically significant.

In conclusion, very few trials investigate the effect of systemic therapy specifically in patients with PM. If anything, especially patients with extensive disease seem to have a poor median OS with systemic therapy. However, there is an important lack of trials in this population addressing recent advances in systemic therapy, such as immunotherapy and targeted therapy.

Intraperitoneal chemotherapy

In this review, 10 trials were included investigating IP chemotherapy in 345 patients with peritoneally metastasized gastric cancer (Table 2).^12–21 Paclitaxel was most frequently used intraperitoneally, followed by docetaxel. In 7 out of 10 trials, IP chemotherapy was repetitively administered, in 3-8 cycles or until unacceptable toxicity occurred. The other 3 trials administered IP chemotherapy only once.^14,16,20 In 9 trials, IP chemotherapy was administered with systemic chemotherapy.^12–20 In 1 trial, a palliative gastrectomy was followed by IP chemotherapy, without the use of systemic chemotherapy.²¹

Table 2.

Overview of trials conducting intraperitoneal chemotherapy.

Author (year)	Country	Design	GCPM n =	Inclusion criteria	Median age (range)	Median PCI (range)	Trial treatment	IP cytostatic	Systemic therapy	AEs (grade ≥3)	Median OS in months	Clinical response rate (complete + partial response)
Ishigami (2018)12	Japan	Phase III RCT	114 vs 50	GCPM, solitary PM	60 (25-74) vs 63 (37-72)	9 vs 4	IP + systemic CTx vs systemic CTx	Paclitaxel	Paclitaxel + S1 vs cisplatin + S1	md	17.7 vs 15.2 (P = .080)	53% (measured in n = 17) vs 60% (measured in n = 5) (P = 1.0)
Li (2021) 13	China	Phase II RCT	40 vs 40	GCPM, solitary PM	56a (±10) vs 58a (±10)	Median between 10-19	IP + systemic CTx (+ gastrectomy) vs systemic CTx	Paclitaxel	Paclitaxel + S1 vs oxaliplatin + S1	37.5% vs 27.5%	13.4 vs 10.8 (P = .011)	52% (measured in all) vs 35% (measured in all)
Shinkai (2018)14	Japan	Phase II, single arm	17	GCPM, solitary PM, no prior treatment	60 (37-71)	Md	IP CTx once + systemic CTx	Paclitaxel	Paclitaxel + cisplatin + S1	0% associated with IP CTx	23.9	80% (measured in n = 5)
Yamaguchi (2013)15	Japan	Phase II, single arm	35	GCPM, solitary PM, irresectable or recurrent GCPM	55 (28-74)	Median between 10-19	IP + systemic CTx	Paclitaxel	Paclitaxel + S1	74%	17.6	71% (measured in n = 7)
Imano (2012)16	Japan	Phase II, single arm	35	GCPM, solitary PM, no prior treatment	66 (32-75)	md	IP CTx once + systemic CTx	Paclitaxel (once)	Paclitaxel + S1	14.4% related to IP CTx; 31.5% related to systemic CTx	21.3	62% (measured in n = 13)
Cho (2017)17	Korea	Phase I/II	Phase I 12 Phase II 39	GCPM, no CTx in last 6 months	Phase I 53 (39-64) Phase II 49 (31-73)	md	IP + systemic CTx	Docetaxel	Capecitabine + cisplatin	md	15.1	60% (measured in n = 5)
Kang (2022)18	Korea	Phase I, dose escalation	15	GCPM, solitary PM, HER2 negative	54a (±12)	20a (±11)	IP + systemic CTx	Paclitaxel	FOLFOX	md	16.6	50% (measured in n = 6)
Lo Dico (2020)19	France	Prospective feasibility trial	6	GCPM with PCI ≥ 15, solitary PM	47 (24-66)	33 (30-39)	IP + systemic CTx	Docetaxel	FOLFOX	33%	13.0	b 67% major macroscopic response (PCI decrease ≥ 25% of initial value)
Imano (2012)20	Japan	Prospective feasibility trial	15	GCPM, solitary PM, no previous systemic CTx	60 (22-75)	Md	IP CTx once + systemic CTx	Paclitaxel	Paclitaxel + S1	13% related to IP CTx; 53% related to systemic CTx	15.8	md
Choi (2011)21	Korea	Phase I, dose escalation	17	GC with resectable primary tumor, PM discovered during operation	50 (37-64)	Md	Palliative gastrectomy + IP CTx	Irinotecan	-	41%	15.6	md

Abbreviations: AEs = adverse events; CTx = chemotherapy; FOLFOX = Folinic acid, Fluorouracil, Oxaliplatin; GC = gastric cancer; GCPM = gastric cancer patients with peritoneal metastases; IP = intraperitoneal; md = missing data; OS = overall survival; PCI = Peritoneal Cancer Index; PM = peritoneal metastases; RCT = randomized controlled trial.

^aMean instead of median

.

^bHistological response instead of response measured via Response Evaluation Criteria in Solid Tumours (RECIST criteria)

.

Trials administering IP in combination with systemic chemotherapy reported a median OS of 13.0-23.9 months.^12–20 No detailed information was reported on the relation to survival and the extent of the peritoneal disease in any of these trials. The radiological response rate ranged from 50% to 80%, measured with the RESIST guideline.^12–27 Lo Dico (2020) documented a PCI decrease of ≥ 25% in 67% of patients during laparoscopic reevaluation.¹⁹

Trials reported an SAE rate of 0%-74% (median 24%). The most frequently reported grade 3-4 adverse events of the systemic chemotherapy with IP chemotherapy consisted of leukopenia, neutropenia, anemia, and anorexia.^12–21 Port-related issues consisted of port infection (3%) and catheter obstruction (3%).¹² Overall, repetitive IP chemotherapy seemed to be well-tolerated; however, the number of cycles differed between studies (median of 3-26 cycles).^{12,13,15,17,19,20} The lowest number of cycles were reported in the trials administering intraperitoneal docetaxel with systemic FOLFOX or systemic capecitabine with cisplatin (median of 3 resp. 8 cycles), while the highest number of cycles were reported in the trials administering systemic paclitaxel or oxaliplatin with S1 and intraperitoneal paclitaxel (median of 11-26 cycles). QoL was not assessed in any of these trials.

In summary, addition of IP chemotherapy to IV chemotherapy suggests a possible benefit in OS benefit.

Hyperthermic intraperitoneal chemotherapy

A total of 9 trials were included performing HIPEC on a total of 265 patients with peritoneally metastasized gastric cancer (Table 3).^22–29 Of these trials, 5 investigated CRS-HIPEC, 1 trial investigated the effect of HIPEC without CRS in elderly patients, and 3 trials investigated CRS-HIPEC+EPIC. Different cytostatics were administered during HIPEC (Table 1).

Table 3.

Overview of trials that conducted hyperthermic intraperitoneal chemotherapy (HIPEC).

Author (year)	Country	Design	GCPM n=	Inclusion criteria	Median age (range)	Median PCI (range)	Systemic therapy	Trial treatment	Cytostatic HIPEC and/or EPIC	Patients with CC0/1 resection (% of patients receiving CRS)	AEs (grade ≥ 3)	Median OS in months (95% CI)
HIPEC (with/without CRS)
Rau (2023)22	Germany	Phase III RCT	52 vs 53	GCPM, solitary PM	56 (48-65) vs 56 (50-63)	Between 7-13 (both arms)	Neoadjuvant and adjuvant CTx	CRS-HIPEC vs CRS	Cisplatin plus mitomycin C	53.8% vs 41.5%	46.3% vs 38.1% peri-operatively (P = .79)	14.9 (97.2% CI 8.7-17.7) vs 6.5 (97.2% CI 7.0-19.4) (P = .047)
Yang (2011)23	China	Phase III RCT	34 vs 34	GCPM, solitary PM	50 (24-75) vs 51 (28-75)	15 vs 15	Adjuvant CTx	CRS-HIPEC vs CRS	Cisplatin plus mitomycin C	20 (58.8%)	14.7% vs 11.7%	11.0 (10-11.9) vs 6.5 (4.8-8.2)b (P = .046)
Topal (2017)24	Belgium	Phase II, single arm	32	GCPM, solitary PM	58 (32-75)	8 (1-20)	Neoadjuvant CTx in 94%	CRS-HIPEC (open HIPEC)	Cisplatin	32 (100%)	28%	16.0 (12.2-24.5)
Chen (2022)25	China	Phase I	22 vs 17	GCPM, age > 65 year, not eligible for CRS	68.6a (±3.6) vs 69.9a (±3.0)	md	Adjuvant CAPOX	HIPEC (closed; 3 cycles within 120 hours)	Lobaplatin plus elemene vs laboplatin alone	-	md	9.4 (5.3-13.5) vs 10.8 (8.5-13.1)
Yang (2009)23	China	Phase I	12	PM in abdominal cancer	53.3 (30-70)	3 (2-18)	md	CRS-HIPEC	HCPT plus Mitomycin C	10 (80%)	0%	md
Mielko (2019)26	Poland	Phase I	30	GCPM, neoadjuvant CTx	55 (28-70)	5a (±4.4)	Neoadjuvant CTx (83% EOX, 17% FLOT)	CRS-HIPEC (with either open or closed HIPEC)	Mitomycin C or oxaliplatin	29 (97%)	46%c	19.3d
CRS-HIPEC + EPIC
Hultman (2013)27	Sweden	Phase II, single arm	18	GCPM, solitary PM	57 (38-74)	12 (5-26)	Neoadjuvant (eg, irinotecan with Nordic FLv, EOX, FLOT)	CRS-HIPEC + EPIC	HIPEC: cisplatin plus doxorubicin; or oxaliplatin alone. EPIC for 5 days with 5-FU or paclitaxel.	7 (88%)	md	10.2 (1.2-34.3)
Hultman (2012)28	Sweden	Phase II, case control study	10 vs 10	GCPM, solitary PM	59 (39-72) vs 62 (40-76)	12 (5-22)	Neoadjuvant, various regimens (eg, Irinotecan with Nordic FLv, EOX, FLOT)	CRS-HIPEC + EPIC vs systemic CTx	HIPEC: cisplatin plus doxorubicin; or oxaliplatin alone. EPIC for 5 days with 5-FU or paclitaxel.	7 (100%)	50% related to CRS-HIPEC vs 50%	15.3 vs 10.4
Kim (2018)29	Korea	Phase I	38	GCPM, age ≤70 years, solitary PM	45.8a (±15.5)	15a	Neoadjuvant CTx in 82%	CRS-HIPEC + EPIC	HIPEC: mitomycin C plus cisplatin. EPIC Day 1: mitomycin C Day 2-5: 5-FU	21 (55.2%)	No AE grade described	19.0

Abbreviations: 5-FU = 5-Fluoruracil; AEs = adverse events; CAPOX = Capecitabine, Oxaliplatin; CC0/1 = completeness of cytoreduction score 0 or 1 (macroscopic complete cytoreduction); CRS = cytoreductive surgery; CRS = Cytoreductive Surgery; CTx = chemotherapy; EOX = Epirubicine, Oxaliplatin, Capecitabine; EPIC = Early Postoperative Intraperitoneal Chemotherapy; FLOT = Docetaxel, Oxaliplatin, Fluorouracil, Leucovorin; GCPM = gastric cancer patients with peritoneal metastases; HCPT = Hydroxycamptothecin; HIPEC = Hyperthermic Intraperitoneal Chemotherapy; md = missing data; Nordic FLv = Nordic 5-Fluorouracil, Leucovorin; OS = overall survival; PCI = Peritoneal Cancer Index; PM = peritoneal metastases; RCT = randomized controlled trial.

^aMean instead of median;.

^bMedian survival from randomization;.

^cClavien-Dindo ≥ 3.

^dMedian survival from day of CRS-HIPEC.

Median OS was 11.0-19.0 months with CRS-HIPEC,^22,24,26 10.0-19.0 months with CRS-HIPEC+EPIC,^27–29 and 9.9 months with HIPEC only in elderly.²⁵ The following factors were associated with a favorable outcome: a low PCI score,^22–25,26,29 complete cytoreduction (CC0),^22–27 a moderately differentiated tumor²⁷ and intestinal type gastric cancer.²⁷ On the contrary, involvement of the small bowel was associated with a poor outcome (P < .001).²⁴ Although a higher PCI was associated with a significantly shorter survival in a phase I trial,²⁹ subgroup analysis in the RCT of Yang (2011) showed a survival benefit in patients treated with CRS-HIPEC compared to treatment with CRS only in patients with a PCI ≥ 20 (n = 14 vs n = 9) (median OS 13.5 months vs 3.0 months, P = .012).²³ On the other hand, patients with a PCI < 20 seemed to have comparable survival with CRS-HIPEC compared to CRS alone (n = 20 vs n = 25) (median OS 10.2 months vs 10.5 months, P = .464).²³ Rau (2023) did not observe a significant benefit on survival with the addition of HIPEC after stratification for PCI.²²

CC0 resection seemed more feasible in trials with patients with a lower median PCI (PCI ≤ 12).^{23–25,28,29} Progression-free survival (PFS) was only described in one trial, reporting a significant improvement of PFS with CRS-HIPEC (7.1 months), compared to CRS alone (3.5 months) (P = .0472).²² Disease-free survival (DFS) was only described in 3 out of 9 trials, reporting a median DFS of 7.8-11.7 months^24,26 in CRS-HIPEC and 9.0 months in CRS-HIPEC + EPIC (in the case of CC0 resection).²⁷

Adverse events (AE) were documented in detail in 5 out of 9 trials only. In the trials performing CRS-HIPEC, the median grade III-IV AE rate ranged from 14.7% to 46.3%, most frequently consisting of infection, obstruction, or wound complications.^22–24,26 A phase III RCT reported no significant difference in severe adverse event (SAE) rate between CRS-HIPEC and HIPEC (14.7% vs 11.7%, P = .839).²³ With CRS-HIPEC plus EPIC, the median grade III-IV AE rate was 50%, and mortality ranged from 5% to 7%.^27–29 A high complication rate was one of the main causes of a high dropout rate in the trials investigating CRS-HIPEC + EPIC.^27–29 Only 2 trials documented described of the QoL, reporting either no treatment effect of CRS/CRS-HIPEC or a better QoL with CRS-HIPEC + EPIC in comparison with systemic chemotherapy alone.^22,28

To conclude, there is possible evidence that CRS and HIPEC when combined with systemic therapy is associated with OS benefit.

Pressurized intraperitoneal chemotherapy

Four trials were included performing PIPAC on a total of 106 patients with peritoneally metastasized gastric cancer (Table 4).^30–33 All trials used a combination of cisplatin and doxorubicin for PIPAC.^30–33 Various systemic treatment regimens were administered prior to PIPAC.^30–32

Table 4.

Overview of trials that conducted pressurized intraperitoneal aerosol chemotherapy (PIPAC).

Author (year)	Country	Design	GCPM n=	Inclusion criteria	Median age (range)	Median PCI (range)	Prior therapy	Trial treatment	PIPAC cytostatic	AEs (grade ≥ 3)	Median OS in months (95% CI)	Clinical response rate (complete + partial response)
Khomyakov (2016)30	Russia	Phase II, open label	31	GCPM, solitary PM, anthracycline/anthracene-dione naïve, age 18-85	52a (25-70)	16 (6-34)	4 courses of XELOX	PIPAC after 4 courses of XELOX	cisplatin + doxorubicin	3.2% (n = 1)	13.0 (md)	Histological response: 60% (in patients with ≥ 2 PIPAC procedures, n = 15)
Struller (2019)31	Germany	Phase II, open label	25	GCPM, solitary PM, disease progression after palliative CTx, measurable target lesion disease on CT	55.1a (±13.0)	15.3a (±10.6)	100% received CTx prior to PIPAC (various regimens) Gastrectomy in 60% (n = 15)	PIPAC	cisplatin + doxorubicin	12% (n = 3)	6.7 (2.5-12.0)	RECIST: 23% (measured in n = 13)
Gockel (2018)32	Germany	Prospective single center registry study	24	GCPM, solitary PM, not eligible for CRS-HIPEC	57 (44-75)	14 (2-36)	83% received first-, second-, or third-line CTx (various regimens)	PIPAC	cisplatin + doxorubicin	Clavien-Dindo: no grade ≥ 3 complications	6.9 (2-20)	md
Kurtz (2018)33	Germany	Prospective single center registry study	26	PM in abdominal cancer	md	md	md	PIPAC	cisplatin + doxorubicin	md	6.8	md

Abbreviations: AEs = adverse events; CRS-HIPEC: cytoreductive surgery and Hyperthermic Intraperitoneal Chemotherapy; CTx = chemotherapy; GCPM = gastric cancer patients with peritoneal metastases; md = missing data; OS = overall survival; PCI = Peritoneal Cancer Index; PIPAC = Pressurized Intraperitoneal Aerosol Chemotherapy; PM = peritoneal metastases; RECIST = Response Evaluation Criteria in Solid Tumors; XELOX = Oxaliplatin, Capecitabine.

^aMean instead of median.

Median OS ranged from 6.0 to 13.0 months calculated from first PIPAC.^30–33 No subgroup analysis was performed on survival with respect to PCI. Khomyakov (2019) reported a 1-year survival rate of 49.8%.³⁰ In patients receiving only one PIPAC procedure, Struller (2019) reported a radiological response rate of 23%.³¹ In patients receiving 2 or more PIPAC procedures, pathological response of 60% was achieved according to the Peritoneal Regression Grading Score (PRGS-score),³⁰ and a decrease in PCI was observed in 4 patients (29%).³¹

Grade III-IV adverse events occurred in 5 out of 106 patients (4.7%).^30–33 Mortality rate was 1%.³³ Struller (2019) reported either abdominal pain or nausea (grade I-V) in all patients receiving PIPAC.³¹ Nevertheless, the treatment appeared not to deteriorate QoL when measured via QLQ-C30 questionnaires.³¹

In conclusion, PIPAC does seem to have a slight OS benefit in comparison with systemic chemotherapy only.

Ongoing trials

On several treatment strategies, trials are being conducted currently. All trials that are still ongoing or yet to be published are summarized in Table 5.^34–44

Table 5.

The ongoing trials or recently completed trails with pending results in patients with peritoneal metastasized gastric cancers.

Author (year)	Country	Treatment modality	Trial name + registration number	Design (intended number of patients)	Treatment of investigation	Inclusion criteria	Outcome measures
On going HIPEC trials
Yang (2017)34	China	CRS-HIPEC vs CRS	NCT03179579	Phase III, RCT (88)	3x HIPEC (paclitaxel + cisplatin + raltitrexed), followed by neoadjuvant CTx (oxaliplatin + S1), CRS, another 3x HIPEC and adjuvant CTx (oxaliplatin + S1) vs neoadjuvant CTx + CRS + adjuvant CTx	GCPM, solitary PM, PCI ≤ 20, no extensive adhesions in peritoneal cavity	mOS, adverse events, risk factors for morbidity and mortality
Koemans (2019)35	The Netherlands	CRS-HIPEC vs systemic CTx	PERISCOPE II trial; NCT03348150	Phase II, RCT (225)	HIPEC: oxaliplatin + docetaxel; systemic: adjuvant CTx vs systemic therapy	GCPM and positive cytology, PCI < 7	mOS, PFS, toxicity, cost and health benefits
Li (2017)36	China	CRS-HIPEC	NCT03023436	Phase III, single arm (220)	CRS-HIPEC with docetaxel + adjuvant cisplatin + fluorouracil	GCPM, solitary PM, PCI < 20,	mOS, PFS, morbidity, mortality, QoL, CTC, ctDNA
Han (2022)37	China	HIPEC + IP CTx vs IP CTx alone	NCT05228743	Phase II/III, RCT (180)	4x HIPEC with PTX + IP/IV PTX + S1 vs NIPS (+ resection)	GCPM, solitary PM, no prior treatment, HER-2 negative	R0 resection rate, 1 year OS
Ongoing IP chemotherapy trials
Lu (2022)38	China	IP + systemic CTx vs systemic CTx	DRAGON-01; ChiCTR-IIR-16009802	Phase III, RCT (238)	IP + systemic CTx vs systemic CTx	GCPM, solitary PM, no prior treatment	mOS, PFS, response rate, conversion surgery rate, adverse events
Yang (2022)39	China	IP + systemic CTx	DRAGON-09; NCT05204173	Phase II, single arm (30)	Sintilimab + IP/systemic PTX + S1	GCPM, solitary PM	1 year survival rate, adverse events, R0 resection rate, 3-year OS, 3-year PFS
INTERACT stomach (2022)40	The Netherlands	IP + systemic CTx	INTERACT stomach; 2021-005907-11	Phase I, dose escalation	IP irinotecan + systemic CAPOX	GCPM, solitary PM, PCI >0	MTD for IP irinotecan, safety, feasibility, pharmacokinetic profile of IP irinotecan
Ramos (2023)41	Brazil	IP + systemic CTx	NCT05541146	Phase II, single arm	IP paclitaxel + systemic CTx (not specified)	GCPM, PCI < 12	mOS, PFS, feasibility
Ongoing PIPAC trials
Luksta (2023)42	Lithuania	PIPAC	NCT05644249	Phase II, single arm	PIPAC: cisplatin + doxorubicin Systemic: FOLFOX	GCPM, solitary PM, HER2 negative	Response after first PIPAC (RECIST), compliance, PCI and regression (PRGS), ascites volume, QoL, OS, PFS, adverse events
Lang (2023)43	Switzerland	PIPAC	NCT04000906	Phase Ib, single arm	PIPAC: cisplatin + nab-paclitaxel	Peritoneal disseminated gastric/pancreatic/ovarian cancer/malignant peritoneal mesothelioma	MTD, adverse events, efficacy (response (PRGS/RECIST)), QoL
Suspended PIPAC trials
Eveno (2019)44	France	Suspended trials	PIPAC ESToK 01; NCT04065139	Phase II RCT (94)	PIPAC (doxorubicin + cisplatin) + systemic CTx vs systemic CTx alone	GCPM, solitary PM, PCI > 8	PFS, OS, safety, tolerability, feasibility of 3 procedures, secondary respectability rate

Abbreviations: CAPOX = Capecitabine, Oxaliplatin; CRS = Cytoreductive Surgery; CTC = circulating tumor cells; ctDNA = circulating tumor DNA; CTx = chemotherapy; FOLFOX = = Folinic acid, Fluorouracil, Oxaliplatin; GCPM = gastric cancer patients with peritoneal metastases; HIPEC = Hyperthermic Intraperitoneal Chemotherapy; IP = intraperitoneal; IV = intravenous; mOS = median overall survival; MTD = maximum tolerated dose; NIPS = neoadjuvant intraperitoneal and systemic therapy; OS = overall survival; PCI = Peritoneal Cancer Index; PFS = progression free survival; PIPAC = Pressurized Intraperitoneal Aerosol Chemotherapy; PM = peritoneal metastases; PRGS = Peritoneal Regression Grading Score; PTX = Paclitaxel; QoL = quality of life; R0 resection = microscopically complete resection; RCT = randomized controlled trial; RECIST = Response Evaluation Criteria in Solid Tumors.

Means instead of median

.

Discussion

In this systematic review, we investigated the effect of all current and new therapeutic options on the survival of patients with peritoneally metastasized gastric cancer. After careful selection, only one trial was included using systemic chemotherapy specifically in patients with peritoneally metastasized gastric cancer, reporting a median OS of merely 6.1-7.3 months.¹¹ It should be noted, however, that this study included patients with a particularly high disease burden, such as massive ascites and/or inadequate oral intake, which are known indicators of poor prognosis.⁴⁵ Interestingly, these outcomes are consistent with Dutch population-based data, which report a median OS of 7.3 months for patients with synchronous peritoneal metastases receiving any form of tumor-directed therapy.⁴ Meanwhile, reported median OS in the broader stage IV gastric cancer population, not limited to peritoneal metastases, typically ranges between 9.0 and 13.0 months.^46–48 However, the highly selective nature of the cohort in the trial of Nakajima et al. (2020), providing chemotherapy without targeted therapy or immunotherapy, limits the generalizability of its findings and may preclude a sound conclusion on the effect of systemic therapy.

In contrast, the reported median OS seems notably longer in the clinical trials performing IP chemotherapy in combination with systemic chemotherapy in patients with peritoneally metastasized gastric cancer, with a median OS of 13.0-23.9 months.^12–21 Nevertheless, it is important to note that only 2 RCTs have evaluated IP chemotherapy in this setting. Of these, Li et al. (2021) demonstrated a statistically significant improvement in survival, while Ishigami et al. (2018) only showed a non-significant trend toward survival benefit. Yet, both RCTs report fairly long survival in the systemic therapy-only cohorts: Li et al. (2021)¹³ reported a median OS of 10.8 months with median PCI between 10-19, and Ishigami et al. (2018)¹² reported a median OS of 15.2 months with median PCI of 4. The remaining survival data are derived from non-randomized cohort studies with considerable heterogeneity, which are subject to limitations, including selection. Still, IP chemotherapy seems feasible with an acceptable toxicity and promising survival.^12–21 Noteworthy, not all trials report specific patient characteristics such as PCI; therefore, the population might be highly selected. Furthermore, immortal time bias has to be taken into account when interpreting these outcomes. The rationale for IP chemotherapy, however, remains strong: delivering chemotherapy directly onto the peritoneum enables longer exposure and a higher drug concentration of the IP cytostatic without additional systemic toxicity.⁴⁹ Different types of cytostatics appear to be both safe and effective. Preferably, large molecular cytostatics are being used for intraperitoneal treatment to minimalize systemic redistribution of the cytostatic agent through the plasma-peritoneum barrier.⁵⁰ Interestingly, several studies report a median OS of 15.8-23.9 months with only a single administration of IP chemotherapy added to systemic chemotherapy.^14,16,20 With the use of an intraperitoneal access port, IP chemotherapy can be easily administered repetitively in combination with systemic chemotherapy in both patients with either low or high PCI scores.^14–21

CRS-HIPEC has been proven beneficial in patients with colorectal cancer and PM with a PCI < 20.⁵¹ The benefit in patients with gastric cancer and PM remains questionable, with high morbidity.^{22–24,26,28} In highly selected patients, HIPEC with or without CRS and/or EPIC show to improve survival in comparison with systemic therapy, with a median OS of 10.0-19.0 months.^22–29 CRS-HIPEC and CRS-HIPEC+EPIC have a comparable median OS, while the addition of EPIC to CRS-HIPEC results in higher dropout rates and complication rates due to toxicity.^22–29 However, the results should be interpreted with caution due to the small sample sizes, heterogenicity of the trials, and the occurrence of selection bias. The heterogenicity relates to use of different intraperitoneal and systemic drugs and different selection criteria, with some trials even including patients with PCI > 20.^22,23,28,29 Since treatment with HIPEC is especially correlated with better survival when a CC0-1 resection is performed, strict selection criteria should be considered for future trials, such as PCI and performance state.^22,23,29

Trials investigating PIPAC yielded less encouraging results. While showing a possible benefit on pathological response rate, this did not translate into an obvious OS benefit in most trials, in comparison with systemic therapy only.^30–33 However, this could be a result of different starting points in OS measurements, due to use of PIPAC as second- or third-line treatment. Furthermore, PIPAC was not associated with deterioration of the QoL and has a relatively low complication rate, however, abdominal pain was very often reported.^31,32 This is in line with presented findings of PIPAC on patients with colorectal cancer and PM.⁵²

This review has several limitations. Firstly, the available studies predominantly exist of phase I and II trails with relatively small cohorts. The comparison and interpretation of the therapeutic effect of the different treatment options needs to be conducted with caution, due to the scarcity of phase III RCT’s, the lack of clearly described systemic therapy backbones, and a great variance in the starting point for measurement of OS makes. Furthermore, this review consists mainly of Asian trials. This is especially the case in the trials investigating IP chemotherapy (90% Asian trials). It is not clear if Asian patients have a similar response and survival to a certain type of treatment in comparison with patients from other ethnic groups. Hence, the results should be interpretated carefully when translated to other ethnic groups. Moreover, it is possible that different treatment modalities are preferable for patients with either limited or extensive PM, as is seen in patients with PM from colorectal origin.⁵¹ Although the extensiveness of PM is an important predictor of prognosis, the extensiveness of PM was not clearly described in most trials.^53,54 Even if it was documented, different scoring systems were used, such as the stage according to the Japanese Gastric Cancer Association, the PCI score, or even separate categories between the PCI score.^45,54 We recommend the incorporation of the PCI score as a stratification factor. Lastly, no trials on immunotherapy or targeted therapy were eligible for inclusion in this review, as none focused specifically on gastric cancer with peritoneal metastases, nor provided subgroup analyses for this population. This represents a significant gap in this field of research. Additionally, most studies therefore lack an adequate control arm, as standard treatment in these trials did not include contemporary systemic therapies such as immunotherapy or targeted agents.

Recently, 2 Dutch trials have been conducted in patients with gastric cancer and PM. First, the PERISCOPE II study was an RCT investigating the role of CRS-HIPEC vs continuation of systemic chemotherapy (after an initial treatment with 3-4 cycles of chemotherapy) in patients with limited GCPM (PCI ≤ 7) or positive cytology.³⁵ The second trial, the INTERACT stomach trial, was a phase I dose-escalation trial to investigate the effect of first line bidirectional chemotherapy in patients with peritoneally metastasized gastric cancer, using intraperitoneal irinotecan combined with systemic CAPOX in patients with gastric cancer and PM of any PCI. We anticipate that the results to be published soon.

In conclusion, only few prospective trials have been conducted specifically in patients with peritoneally metastasized gastric cancer. While intraperitoneal chemotherapy and CRS-HIPEC have shown promising results in selected cohorts, evidence from randomized trials remains limited. Further prospective trials are urgently needed to clarify the role of these treatment modalities alongside contemporary systemic therapies, including immunotherapy and targeted therapy. To enhance the generalizability of the outcomes of future trials, we advocate for the use of RCTs to substantiate evidence, for a clear description of a specific systemic therapy backbone, and for both documentation and stratification of the PCI score with the interpretation of the trial results.

Author contributions

Marion W. Tops-Welten (Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Project administration, Validation, Visualization, Writing—original draft, Writing—review & editing), Laskarina J.K. Galanos (Data curation, Investigation, Validation, Writing—review & editing), Geert-Jan Creemers (Conceptualization, Supervision, Writing—review & editing), Misha D.P. Luyer (Conceptualization, Supervision, ­Writing—review & editing), Ignace H.J.T. De Hingh (Conceptualization, Supervision, Writing—review & editing), and Irene E.G. van Hellemond (Conceptualization, Data curation, Investigation, Methodology, Supervision, Writing—review & editing)

Supplementary material

Supplementary material is available at The Oncologist online.

Funding

Misha D.P. Luyer received research funding from Galvani and consultancy fees from Medtronic. Ignace H.J.T. De Hingh reports an unrestricted research grant from RanD/QPS, Roche, and the Hanarth Fonds, outside the submitted work, paid to the institution.

Conflicts of interest

None declared.

Data availability

All data addressed in this review were obtained from publicly available and published data sources. Within the search strategy, the following databases were used: PubMed, Embase, Cochrane Library, and ClinicalTrials.gov. To facilitate reproducibility, the full search strategy is available as supplementary material with this article.