The Development of Peritoneal Metastasis from Gastric Cancer and Rationale of Treatment According to the Mechanism

In the present article, we describe the normal structure of the peritoneum and review the mechanisms of peritoneal metastasis (PM) from gastric cancer (GC). The structure of the peritoneum was studied by a double-enzyme staining method using alkaline-phosphatase and 5′-nucreotidase, scanning electron microscopy, and immunohistological methods. The fundamental structure consists of three layers, mesothelial cells and a basement membrane (layer 1), macula cribriformis (MC) (layer 2), and submesothelial connective tissue containing blood vessels and initial lymphatic vessels, attached to holes in the MC (layer 3). Macro molecules and macrophages migrate from mesothelial stomata to the initial lymphatic vessels through holes in the MC. These structures are characteristically found in the diaphragm, omentum, paracolic gutter, pelvic peritoneum, and falciform ligament. The first step of PM is spillage of cancer cells (peritoneal free cancer cells; PFCCs) into the peritoneal cavity from the serosal surface of the primary tumor or cancer cell contamination from lymphatic and blood vessels torn during surgical procedures. After PFCCs adhere to the peritoneal surface, PMs form by three processes, i.e., (1) trans-mesothelial metastasis, (2) trans-lymphatic metastasis, and (3) superficial growing metastasis. Because the intraperitoneal (IP) dose intensity is significantly higher when generated by IP chemotherapy than by systemic chemotherapy, IP chemotherapy has a great role in the treatment of PFCCs, superficial growing metastasis, trans-lymphatic metastasis and in the early stages of trans-mesothelial metastasis. However, an established trans-mesothelial metastasis has its own interstitial tissue and vasculature which generate high interstitial pressure. Accordingly, it is reasonable to treat established trans-mesothelial metastasis by bidirectional chemotherapy from both IP and systemic chemotherapy.


Introduction
Gastric cancer (GC) is the second leading cause of death from cancer. Peritoneal metastasis (PM) is the most common form of metastasis in GC, and PM is about 14% of primary GC cases [1,2]. However, patients have a median survival time of 3-6 months [1,2]. Until the early 1990s, GC with PM was considered an incurable disease because it could not be cured by surgery or systemic chemotherapy alone [3,4]. Even after complete resection of a small number of PMs with gastrectomy plus lymph adenectomy, residual micrometastasis on the peritoneal surface always proliferates and almost all patients will die. Systemic chemotherapy using modern drugs has limited effects on PM [3,4] because only small amounts of systemically administered drugs can enter the peritoneal cavity, and even effective regimens are inevitably interrupted due to the development of side effects or regrowth of multidrug-resistant cancer cells.
In the late 1990s, a combination of cytoreductive surgery (CRS) and perioperative chemotherapy for the treatment of PM was proposed as a comprehensive treatment by the Peritoneal Surface Oncology Group International (PSOGI) [5]. In CRS, all the macroscopically detectable tumors are removed by D2 gastrectomy and peritonectomy [5,6]. However, even macroscopic complete cytoreduction leaves invisible micrometastases in most cases [7]. To eradicate the micrometastases before and after CRS, neoadjuvant IP chemotherapy and intraoperative hyperthermic intraperitoneal chemoperfusion (HIPEC) were developed [8][9][10][11]. The comprehensive treatment improved the long-term survival of GC patients with PM and 10% of patients survived longer than 10 years [11]. Accordingly, the treatment is now considered a curative approach [6,9,11]. For the development of more effective treatments to improve survival, it is important to clarify the mechanisms of the formation of PM.
The present chapter presents the mechanisms of PM from GC, and the rationale for eliminating micrometastasis by chemotherapy.
The peritoneal cavity normally has 40 to 50 mL of ascites that always circulate through the submesothelial lymphatic vascular system, also known as the peritoneal systemic circulation. Ascites and the glycocalyx such as hyaluronic acid produced from mesothelial cells play a role as a lubricant for the viscera to slide over the visceral and parietal peritoneal surface without friction. Ascites play a role as a medium through which peritoneal macrophages pass to patrol the peritoneal microenvironment [12][13][14]. Figure 1 shows the fundamental structure of the peritoneum. Mesothelial cells cover the basement membrane (BM), which in turn covers a layer of the collagen plate known as the macula cribriformis (MC). Lymphatic vessels permeate the shallow subperitoneal space in Morrison's pouch (Figures 2A-C and 3), the falciform ligament ( Figure 4A-C), the pelvic peritoneum ( Figure 5) and the para-colic gutter ( Figure 6) [12]. The blood-peritoneal barrier (BPB) consists of mesothelial cells, BM, MC, and connective tissues between mesothelial cells and submesothelial blood vessels ( Figure 1). Some submesothelial lymphatic vessels usually look like a blind loop (Figure 4), and are attached to the small holes in the MC and mesothelial cell gaps. These mesothelial cell gaps are named lymphatic stomata (Figure 1, right) [14]. Peritoneal fluid-containing electrolytes are absorbed from the subperitoneal lymphatic vascular system, and macromolecules and inflammatory cells are resorbed and migrate through lymphatic stomata [14,15]. Alkaline-phosphatase and 5 -nucleotidase (5 -Nase) double enzyme staining differentiates subperitoneal blood capillaries (Figure 2A, blue) from lymphatic vessels (brown) (Figure 2A). Below the submesothelial BM, multiple holes ( Figure 7) are found in the MC, and the blind loops of the submesothelial lymphatic vessels are attached to the holes (Figures 1-8). Activated carbon (CH44) is present in the blind loop of the submesothelial lymphatic vessels (also known as initial lymphatic vessels, which are stained by 5 -Nase enzyme staining (Figure 8)) 2 days after IP injection (Figure 8, left). CH44 is absorbed between the lymphatic endothelial cells (Figure 8, right) and is in the subperitoneal lymphatic vessels in Morrison's pouch (5 -Nase enzyme staining, Figure 3). These results indicate that intraperitoneally injected CH44 is absorbed through initial lymphatic vessels [14]. These lymphatic vascular structures are found in Morrison's pouch (Figures 3 and 7), the pelvic peritoneum ( Figure 5) and falciform ligament ( Figure 4) and have a three-layered fundamental anatomical structure ( Figure 6) that includes a first layer consisting of mesothelial cells and BM, a second layer known as the MC, and a third layer containing the initial lymphatic vessels that are attached to holes in the MC. Macro molecules and macrophages migrate from mesothelial stomata and move into the initial lymphatic vessels through holes in the MC ( Figure 6). . The blood-peritoneal barrier (BPB) consists of mesothelial cells, BM, MC, and connective tissues between the MC and submesothelial blood vessels. Some submesothelial lymphatic vessels are attached to small holes in the MC and mesothelial cell layer (left), *, which are named lymphatic stomata (right). Scanning electron microscopic study was performed by Miura M, one of the authors of the article, and the techniques of SEM are described in reference [16].       The lymphatic vascular system of the diaphragm is slightly different from that of the pelvis and Morrison's pouch. On the abdominal surface of the diaphragm, the openings in the mesothelium that connect it with the submesothelial lymphatic vessels (initial lymphatic vessels) are defined as lymphatic stomata, whereas unattached mesothelial gaps to the initial lymphatic vessels are not defined as lymphatic stomata ( Figure 9C). Negative pressure due to inspiration helps to absorb peritoneal fluid and macro molecules through the lymphatic stomata of the diaphragm, and the absorbed materials migrate in the lymphatic vessels, which vertically run through diaphragmatic muscle ( Figure 9A). Blood vessels below the mesothelial cells have a role in the absorption of peritoneal fluid ( Figure 9B). As shown in Figure 10, holes in the diaphragmatic MC connect with the initial lymphatic vessels ( Figure 10).  The greater and lesser omentum have specialized structures know as omental milky spots (OMS), which are associated with inflammatory cell migration and ascite absorption. The OMS are small organs 15 to 800 m in diameter, and their mean number is 35/cm 2 in infants and 2/cm 2 in adults [15,17]. The outer surface of OMS is covered with cuboidal mesothelial cells punctured by stomata and supported by a BM (Figures 11 and 12), and mesothelial cells surrounding OMS are flat ( Figure 12, left). The MC, glomerular blood vessels and initial lymphatic vessels are found in the second and third layers ( Figure 11). The fundamental structure of OMS is similar to that of diaphragm/falciform ligaments and the peritoneum of the pelvis, para-colic gutter and Morrison's pouch ( Figure 6). The number of OMS change with age and peritoneal inflammatory status ( Figure 11) [17].

Mechanisms of the Formation of PM from GC
PM is characterized by a multi-step process consisting of (1) detachment of cancer cells from the serosal surface of the primary GC, (2) migration of peritoneal free cancer cells (PFCCs) through ascites on the distant peritoneal surface and attachment of PFCCs on the peritoneal surface by adhesion molecules, (3) proliferation of PFCCs on the peritoneal surface or invasion of PFCCs into the subperitoneal tissue through the concerted actions of matrix-digesting enzymes, adhesion molecules and motility factors expressed from cancer cells and stromal cells, and (4) neoplastic proliferation accompanying immature stromal elements and angiogenesis [18].
In 1997, Sugarbaker reported two patterns in PM formation concepts, i.e., randomly proximal distribution and redistribution pattern [19].

Mechanisms of the Cell Spillage into the Peritoneal Cavity
Spillage of cancer cells into the peritoneal cavity occurs from the serosal surface of the primary tumor or cancer cell contamination from lymphatic and blood vessels torn during surgical procedures.
In the process of cancer cell detachment from the primary tumor, dysfunction of homophilic cell-cell adhesion molecules plays an important role. Analyses of tight and adherence junction molecule expression demonstrate that a reduction in the expression of claudin, occludin and E-cadherin induces cell dispersion [20][21][22][23] and is frequently found in poorly differentiated adenocarcinomas of the stomach [24][25][26]. In GC, poorly differentiated adenocarcinomas have a significantly higher PM potential than the differentiated type and are characterized by downregulation of E-cadherin and its catenin partner [25][26][27][28]. E-cadherin expressed on the adherence junction plays a major role in cell-cell adhesion. The loss of cadherin induces the loosening of the adhesion, resulting in the dispersion of cells.

Adhesion of PFCCs to the Distant Peritoneum
PFCCs migrate through ascites and reach the distant peritoneum ( Figure 13). While rolling on mesothelial cells, PFCCs adhere to mesothelial cells via adhesion molecules and their ligands. In homophilic heterotypic adhesion, P-cadherin participates in the loose attachment of PFCCs to mesothelial cells [29]. Integrin L2 and 2 expressed on PFCCs heterotopically bind with PCAM-1 and VCAM-1 from mesothelial cells [14,30].
Hyaluronic acids from mesothelial cells are specific ligands of the transmembrane glycoprotein CD44 from PFCCs, and CD44 isoforms enhance local growth and metastasis of cancer cells [31,32].
The Syalyl Lewis A antigen has the carbohydrate structure on cancer cells needed for E-selectinbinding to mesotherial cells [33].
CA 125 is a mucin-like glycoprotein that is upregulated in some cancer cells [34,35], whereas mesothelin is expressed by normal mesothelial cells and binds with CA125 [36].
Recently, Arita T et al. reported that tumor-derived exosomes (TEX) may play a critical role in the development of PM, which may be due to inducing increased expression of adhesion molecules from mesothelial cells [37]. TEX are known to contain mRNA and micro RNA from GC cells, and these may trigger mesothelial cell transformation, leading to a cancer-preferable microenvironment.

Mechanisms of Trans-Mesothelial Metastasis
After PFCCs' attachment to mesothelial cells, PFCCs digest the BM exposed between shrunk mesothelial cells. Akedo H et al. observed three growth patterns of cancer cells when rat hepatoma cell monolayers were co-cultured with a rat mesothelial cell monolayer [41]. Tumor cells either formed "pile-up" nests on the mesothelial monolayer, exhibited invasive growth between adjacent mesothelial cells, or failed to attach and grew in suspension. Figure 16 shows a micrometastasis of a differentiated adenocarcinoma growing on the small bowel mesentery. Cancer cells show Ki 67 immunoreaction, suggesting high proliferative activity ( Figure 16B). CD34-positive stromal cells are observed at the invasion front of the micrometastasis ( Figure 16C). CD34 is a transmembrane glycoprotein that is expressed on the stem cells and fibroblasts of immature tissues [49]. As shown in Figure 16D, no newly formed blood vessels are detected in the stroma of micrometastasis. In contrast, Figure 17 shows the histological structure of an established metastasis from a differentiated adenocarcinoma. The cancer cells show high proliferative activity ( Figure 17B), and the stromal cells are CD34-positive stromal cells ( Figure 17C), suggesting the immaturity of the interstitial tissue. Newly formed blood vessels are found in the stroma of established PM ( Figure 17D). These results indicate that PFCCs attach to the peritoneal surface and invade the subperitoneal tissue by a mechanism involving the concerted expression of motility factors [50][51][52][53][54][55][56] matrix digesting enzymes [57][58][59][60][61][62][63][64][65][66][67][68][69], and adhesion molecules [44][45][46]70,71]. As PFCCs grow in the subperitoneal tissue, a new vascular network forms in response to the secretion of angiogenesis factors from cancer cells and surrounding interstitial cells (Figure 17) [72][73][74]. As shown in Figure 17C,D, CD34-positive interstitial fibroblasts are found in the stromal tissue, and newly formed blood vessels are detected in the stromal tissues of established trans-mesothelial metastasis.  Figure 18 shows the micrometastasis of a poorly differentiated adenocarcinoma that formed in the rectal Douglas pouch. Cancer cells invading from the peritoneal surface are found in the subserosal layer and invade into the rectal muscle layer ( Figure 18A) and submucosal layer ( Figure 18B). These cancer cells are immunoreactive with Ki67 monoclonal antibody, suggesting high proliferative activity ( Figure 18C). CD34-positive cells are detected in the stromal tissue ( Figure 18D), indicating that the poorly differentiated adenocarcinoma easily invades through the gaps in the muscle layer by a mechanism that uses motility factors, adhesion molecules and matrix digesting enzymes. Metastasis will become established upon cancer cell proliferation with angiogenesis induced by CD34positive stromal cells. This type of peritoneal metastasis is named trans-mesothelial metastasis.

Mechanisms of Superficial Growing Metastasis
Cancer cells located greater than 100 m from blood vessels will die due to an insufficient supply of oxygen [18,42,75]. The BPB occupies the space between the mesothelial cell surface and submesothelial blood vessels (Figure 1). In Morrison's pouch, the falciform ligament, and diaphragmatic peritoneum, there are subperitoneal vascular networks (a short-distance BPB), located just beneath the MC (Figures 2 and 9) [12] and PFCCs with low invasive activity proliferate by absorbing oxygen from the subperitoneal blood vessels on the peritoneal surface with a short-distance BPB [12]. This type of metastasis is named the superficial growing metastasis (Figure 19) [14]. In contrast, the peritoneum on the anterior abdominal wall ( Figure 20A) has few submesothelial blood vessels ( Figure 20C). The peritoneum covering the anterior abdominal wall is considered to be resistant to the development of superficial growing metastasis because of its long-distance BPB ( Figure 20).

Mechnisms of Trans-Lymphatic Metastasis
In trans-lymphatic metastasis, PFCCs migrate into the initial lymphatic vessels. Cytokine secretion from the PFCCs causes mesothelial cell shrinkage ( Figure 14) and hole formation in the MC ( Figure 21A), thereby exposing the initial lymphatic vessel (Figures 6, 7, 21 and 22), and resulting in a direct route from the peritoneal cavity to the initial lymphatic vessels (Figures 6 and 22A). Scanning electron micrography shows that the PFCCs migrate into the initial lymphatic vessels through mesothelial lymphatic stoma and holes in the MC ( Figure 21A-C).  Schematic diagram of trans-lymphatic metastasis is shown in Figure 22A. PFCCs invade the initial lymphatic vessels and proliferate in the lymphatic lacunae ( Figure 22B), and their growth finally destroys the lymphatic vessels ( Figure 22C). Trans-lymphatic metastasis develops in the initial lymphatic vessel-rich sectors in the parietal peritoneum, showing a pink area ( Figure 23). As shown in Figure 20B, initial lymphatic vessels are not found in anterior abdominal wall, and trans-lymphatic metastasis is not found in the sector.
OMS are common sites of trans-lymphatic metastasis [12][13][14][15]76]. PFCCs are adsorbed to the stomata between the cuboidal mesothelial cells covering the OMS and invade the omental lymphatic vessels through holes in the MC (Figures 11 and 12).
On the small bowel mesentery, 2-3 cm in from the attachment sites on the small bowel, milky spot-like structures are found ( Figure 24). PM from GC is frequently found in the peritoneal area ( Figure 24A), and intraperitoneally injected CH44 is adsorbed in the same area of the small bowel mesentery ( Figure 24B), suggesting the absorption of CH44 by the initial lymphatic vessels in the milky spots of small bowel mesentery. Scanning electron micrography shows the oval-shaped structure of the milky spot-like structure covered by cuboidal mesothelial cells ( Figure 24C). Below the cuboidal mesothelial cells, MC with holes is detected ( Figure 24D).  Scanning electron micrography shows the omental milky spots (oval-shaped structures) covered with cuboidal mesothelial cells. Between the cuboidal mesothelial cells, stomata-like holes are observed. Initial lymphatic vessels can be seen in the epiploic appendage of the colon (Figure 25A), and metastasis on the epiploic appendage should be removed ( Figure 25B,C).   Figure 26 shows the omental bursa consisting of the omental sac ( Figure 26A,B), superior recess ( Figure 26C), and anterior vestibule ( Figure 26D). PFCCs migrate through the foramen of Winslow into the omental bursa ( Figure 26). Around the duodenojejunal junction, there are two pockets, i.e., the superior and inferior recess (Figure 27). Figure 27C shows a metastasis on the inferior recess. In the left side of sigmoid mesocolon, there is an intersigmoid recess ( Figure 28). The number of recesses varies from case to case ( Figure 28B,C). Figure 28 C shows the metastases on the intersigmoid recess. The pelvic cavity has many pockets. Males have a rectovesical pouch ( Figure 29A), and females have the Douglas pouch and vesicouterine pouch ( Figure 29B). Figure 29C is in intraperitoneal view of the en-bloc resection pelvic pockets, and Figure 29D shows the resected specimen of the rectum, uterus and pelvic peritoneum.

Selection of the Route of Chemotherapy with Special Reference to the Mechanisms of PM formation
Systemic chemotherapy has little effect on PM because the intraperitoneal (IP) transport of chemotherapeutic agents is limited [77]. BPB with average distance of 90 µm hinders the movement of drugs from the subperitoneal blood vessels to the peritoneal cavity [77]. In contrast, IP chemotherapy could generate significantly higher intraperitoneal dose intensity than systemic chemotherapy (Table 1) [78]. After IP administration, the drug penetration distances into the subperitoneal tissue are different from drug to drug (Table 1). In general, the larger the molecular weight of the drug, the longer the drug stays in the peritoneal cavity. Drugs with higher molecular weights are recommended for IP chemotherapy to treat PFCCs, superficial growing metastasis (Figure 19), trans-lymphatic metastasis ( Figure 22) and the early stage of trans-mesothelial metastasis (Figure 16), where the cancer cell is growing in the superficial submesothelial layer without angiogenesis ( Figure 16). Table 1. Drug molecular weights, pAUC/sAUC, penetration distance into the subperitoneal tissue after intraperitoneal administration, maximum tolerate doses (MTD) and thermal enhancement with each drug. NA: not analyzed, pAUC: area under the curve in peritoneal cavity, sAUC: area under the curve in serum [10,12]. Taxans show anti-cancer effects to injure the function of the spindle body by promoting microtubule polymerization. Because taxol and docetaxel have hydrophobic properties, these drugs are conjugated with cremophol EL and polysorbate 80 to be water-soluble components [78]. Taxol and docetaxel are gradually released from the carriers to ascites after intraperitoneal administration. Accordingly, these drugs tend to stay for a long time after intraperitoneal administration. The concentrations of these drugs in ascites are maintained for 24 h at significantly high levels to kill cancer cells [78]. Accordingly, these drugs may be effective in the treatment of PFCCs, superficial growing metastasis, and translymphatic metastasis (Table 1) [78,79]. For the treatment of trans-mesothelial metastasis, intraperitoneal administration of drugs capable of deeper penetration into the subperitoneal tissue is recommended (Table 1). After IP administration, cisplatin, oxaliplatin, and carboplatin can penetrate a depth greater than 1 mm [80,81]. However, an established trans-mesothelial metastasis has its own interstitial tissue and vasculature, resulting in high interstitial pressure in the metastasis (Figures 17 and 18) [42]. Accordingly, drugs that penetrate a metastasis with high interstitial pressure are easily excreted into the normal tissue with low interstitial pressure. A bidirectional chemotherapy from both sides of the intraperitoneal and systemic administration is a reasonable option for treating established trans-mesothelial metastasis ( Figure 17) [7,8,10]. Metastases invading the submucosal layer and proper muscle of bowel should be treated with systemic chemotherapy (Figure 18).

Drugs
Because chemotherapeutic drugs can injure proliferating cancer cells but penetrate the peritoneal surface only a short distance, IP chemotherapy should be repeated until drug resistance appears or macroscopic complete cytoreduction can be performed.

Effects of Neoadjuvant Intraperitoneal and Systemic Chemotherapy (NIPS) on Lymph Node Metastasis
IP chemotherapy is considered an effective method of treating PM from GC [7,8]. Yonemura Y et al. developed a novel neoadjuvant chemotherapy combining intraperitoneal and systemic chemotherapy, which is named NIPS [7,8,11]. In NIPS, intraperitoneal administrations of 40 mg of docetaxel and cisplatinum with 500 mL of normal saline are performed on day 1 and 14, and oral intake of 60 mg/m 2 of S1 starts from day 1 to day 14. After 1 week rest (from day 15 to day 21), NIPS repeats for at least 3 cycles. One month after the last cycle of NIPS, patients opted for performing cytoreductive surgery and received a gastrectomy plus D2 lymph adenectomy and peritonectomy. They studied the effects of lymph node metastasis (LNM) after NIPS and compared the results to the no NIPS group [82].
The incidence of N0 cases was significantly higher in the NIPS group (37/107; 34.6% vs. 14/136; 10.3%) (p < 0.0001). Survival was significantly longer after NIPS plus cytoreductive surgery than in the non-NIPS group. NIPS is a very effective method of controlling LNM from GC. After IP administration of chemotherapeutic drugs, extremely higher concentrations of chemotherapeutic drug are absorbed through OMS and the efferent lymphatic fluid drains into the regional lymph nodes of the stomach. As a result, LNM from GC is exposed to much higher concentrations of chemotherapeutic drugs than when chemotherapy is given systemically. This feature of the lymphatic circulation accounts for the much greater effects of NIPS on LNM.

Effects of NIPS on PM from GC
Effects of NIPS on PM from GC has been reported from Japanese surgical oncologists. This trial failed to show statistical superiority of survival in the NIPS group as compared with the control group. However, the response rate analyses suggested possible clinical benefits of NIPS for PM from GC [83].
Yonemura Y et al. studied the effects of NIPS by laparoscopy, and the peritoneal cancer index (PCI) was significantly reduced after three cycles of NIPS. These results indicate that NIPS is effective in PCI reduction [84]. The survival of GC patients with PM treated by NIPS plus CRS was significantly better than those treated with cytoreduction without NIPS [12]. Additionally, Yonemura et al. reported that post-NIPS PCI ≤6 was the strongest independent prognostic factor [10,11]. Accordingly, NIPS is essential to improve postoperative survival of GC patients with PM by reducing PCI and micrometastasis left on the preserved peritoneal surface after CRS.

Effects of HIPEC on PM from GC
Yonemura et al. studied the effect of laparoscopic HIPEC (LHIPEC) on PM from GC. LHIPEC receiving cisplatin 50 mg and docetaxel 40 mg in 4000 mL of normal saline at 43 ± 0.5 • C for 60 min were performed in 55 GC patients with PM, and laparoscopy was again performed one month later. PCI was significantly reduced and the positive cytology became negative in 62% of patients with positive cytology at the time of LHIPEC [84].
Brandl A collected and analyzed a worldwide cohort of patients treated with cytoreductive surgery and HIPEC with long-term survival in order to explore relevant patient characteristics [85]. From an analysis of 448 patients, a total of 28 patients with a mean PCI of 3.3 survived longer than 5 years after CRS plus HIPEC. The overall median survival was 11.0 years (min 5.0; max 27.9). The predictor completeness of cytoreduction (CC-0) and PCI ≤ 6 were present in 22/28 patients. They concluded that the completeness of cytoreduction and low PCI seemed to be crucial and that long-term survival and even cures are possible in patients with PM of GC treated with CRS and HIPEC Yang XJ performed a randomized phase-III study to evaluate the efficacy and safety of CRS plus HIPEC for PM from GC [86]. Sixty-eight gastric PC patients were randomized into CRS alone (n = 34) or CRS plus HIPEC (n = 34) receiving cisplatin 120 mg and mitomycin C 30 mg each in 6000 mL of normal saline at 43 ± 0.5 • C for 60-90 min. The median survival was 6.5 months in CRS and 11.0 months in the CRS plus HIPEC groups (p = 0.046). Multivariate analysis found that CRS plus HIPEC, synchronous PC, CC 0-1, systemic chemotherapy ≥ 6 cycles, and no serious adverse events were independent predictors for better survival. They concluded that CRS + HIPEC with mitomycin C 30 mg and cisplatin 120 mg may improve survival with acceptable morbidity [86].
These results indicate that HIPEC is directly effective on PM from GC, and HIPEC may improve GC patients' survival after complete resection of the PM.

Future Perspectives
IP immunotherapy offers a novel approach for the control of regional disease of the peritoneal cavity by breaking immune tolerance. These strategies include heightening Tcells and vaccine induction of anti-cancer memory against tumor-associated antigens [87].
Catumaxomab, a non-humanized chimeric antibody, is characterized by its unique ability to bind to three different types of cells: tumor cells expressing the epithelial cell adhesion molecule (EpCAM), T lymphocytes (CD3) and also accessory cells (Fcy receptor). Because up to 90% of gastric cancers express EpCAM, IP infusion of catumaxomab after complete resection of all macroscopic disease could therefore efficiently treat macroscopic residual disease [88]. Knödler M performed a prospective randomized phase-II study investigating the efficacy of catumaxomab followed by chemotherapy (arm A, 5-fluorouracil, leucovorin, oxaliplatin, docetaxel, FLOT) or FLOT alone (arm B) in patients with GC and PM. However, no survival benefit was found in the treatment group [88].
Nivolumab, an anti-programmed cell death-1 (PD-1) antibody, has been developed and survival benefit was obtained in GC patients with PM [89]. The combination of CRS plus HIPEC and postoperative nivolumab may improve postoperative survival, but no randomized phase-III study was reported.
Nab-paclitaxe (nanoparticle alubumin-bound paclitaxel) has effective transferability to tumor tissues and strong antitumor effects for peritoneal metastasis [90]. Ishikawa M reported that nanoparticle alubumin-bound (nab)-PTX treatment has beneficial effects on the survival of GC patients with PM as compared the RAM plus solvent-based (sb) paclitaxel [90].
Pressurized intraperitoneal aerosol chemotherapy (PIPAC) was introduced as a new treatment for patients with peritoneal metastases in 2011. Adverse events greater than grade 2 occurred after 12-15% of procedures. An objective clinical response of 50-91% was found for gastric cancer (median survival of 8-15 months). Alyami M et al. reported that PIPAC was safe, and the objective response and quality of life were encouraging. Therefore, PIPAC can be considered as a treatment option for refractory, isolated peritoneal metastasis [91]. However, no long-term survival after PIPAC has been reported.
Patients with PM may be improved by gene therapy [92]. Many ideas have been reported, but large-scale clinical studies have not yet been performed.
We are awaiting the development of new effective nanomolecules, cancer-specific antibodies or anti-cancer drugs for the combination of treatment options with CRS.
In the surgical field, further studies are needed in order to improve existing selection criteria for cytoreductive surgery.

Conclusions
The fundamental structure of the peritoneum consists of three layers: mesothelial cells and the basement membrane (layer 1), the macula cribriformis (MC) (layer 2), and the submesothelial connective tissue containing blood vessels and initial lymphatic vessels (layer 3). Macro molecules and macrophages migrate from mesothelial stomata to the initial lymphatic vessels through holes in the MC.
These structures are characteristically found in the diaphragm, omentum, paracolic gutter, pelvic peritoneum, and falciform ligament.
The first step of PM is the spillage of cancer cells (peritoneal free cancer cells; PFCCs) into the peritoneal cavity from the serosal surface of the primary tumor or cancer cell contamination from lymphatic and blood vessels torn during surgical procedures.
After PFCCs adhere to the peritoneal surface, PMs form by three processes, i.e., (1) trans-mesothelial metastasis, (2) trans-lymphatic metastasis, and (3) superficial growing metastasis. Because the intraperitoneal (IP) dose intensity is significantly higher when generated by IP chemotherapy than by systemic chemotherapy, IP chemotherapy plays a great role in the treatment of PFCCs, superficial growing metastasis, trans-lymphatic metastasis and early stage of trans-mesothelial metastasis. However, an established transmesothelial metastasis has its own interstitial tissue and vasculature, which generates high interstitial pressure.
Accordingly, it is reasonable to treat established trans-mesothelial metastasis by bidirectional chemotherapy from both IP and systemic chemotherapy.  Institutional Review Board Statement: The study was approved by the ethical committee of Kishiwada Tokusyukai Hospital, the protocol code of 19-35, that title is "A study of comprehensive treatment for peritoneal metastasis". It was approved on 15 October 2007.
Informed Consent Statement: Informed consent was obtained from all subjects involved in the study. Written informed consent has been obtained from the patients to publish this article.