Dissemination or spread of malignant tumor throughout the peritoneal (abdominal and pelvic) cavity.
Transcoelomic (meaning “across the peritoneal cavity”) metastasis refers to the dissemination of malignant tumors throughout the surfaces and organs of the abdominal and pelvic cavity covered by peritoneum. Transcoelomic metastasis can occur as a result of invasion into the peritoneal cavity by (1) a primary cancer arising from within the abdominal/pelvic cavity, e.g., ovarian cancer; (2) as a manifestation of systemic metastasis following haematogenous or lymphatic invasion by a primary cancer, e.g., advanced breast cancer; or (3) following intraperitoneal seeding during surgical manipulation, e.g., during surgical resection of a colorectal tumor.
The incidence of transcoelomic metastasis is higher with tumors that arise from the peritoneal cavity, e.g., ovarian (up to 70% of patients at presentation) and colorectal (up to 28% of patients at presentation). In contrast, extraperitoneal cancers, e.g., breast and lung, are associated with a much lower overall incidence of transcoelomic metastasis, although certain histological subtypes, e.g., infiltrating lobular breast cancers, have demonstrated a greater predilection for metastases to the gastrointestinal tract, gynecological organs, and peritoneum/retroperitoneum. This suggests that while the location of the primary tumor may be a key determinant in the development of transcoelomic metastasis, the tumor phenotype is also an important factor. Hence it appears that a combination of anatomical and tumor-specific factors is involved in the transcoelomic metastatic process.
Transcoelomic metastases contribute considerably to the morbidity associated with carcinomatosis because they have the capacity to affect multiple vital organs within the abdomen. Common examples include bowel obstruction caused by lesions along the gastrointestinal tract, and renal failure caused by obstruction of the ureters. In addition, transcoelomic metastases are frequently associated with the formation of malignant Aurora A, resulting in raised intra-abdominal pressure with consequent abdominal distension and discomfort. This results in early satiety, leading to dietary deficiency, impaired circulation of blood and lymphatic vessels, and respiratory compromise secondary to diaphragmatic splinting. Hence, there are potentially significant therapeutic advantages to be gained in understanding the process of transcoelomic metastasis.
Mechanisms of Transcoelomic Metastasis
Models of Metastasis
Two models have been hypothesized for the genetic origins of tumor metastases. The first model, often referred to as the seed-and-soil hypothesis, is that tumors are genetically heterogeneous and metastases arise from clones with a genetically acquired metastatic phenotype, which determines the final site of metastasis. The alternative hypothesis, the stochastic model, is that metastatic cells do not represent a genetically selected clone distinct from the primary tumor, but arise as a stochastic event from tumor cell clones genetically identical to the primary tumor. Recent studies exploring this question using in vivo models have suggested a combination of both models of metastasis. Regardless of metastatic model, there are certain observed characteristics that appear to be important for transcoelomic metastastic progression, in which complex cellular adaptations need to occur after cell detachment from the primary tumor mass to ensure survival within the peritoneal cavity.
Anchorage-independent growth and the ability to resist anoikis is a vital step for the initiation of metastasis. This process appears to involve the increased expression of survivin and X-linked inhibitor of apoptosis (XIAP), members of the inhibitor of apoptosis protein (IAP) family, which suppress apoptosis by inhibition of Caspases. Other mediators of anoikis resistance include the family of extracellular matrix (ECM) to cell-adhesion molecules known as Integrins. Alterations in levels of integrin-mediated ECM-ligand binding have been found in many different tumor types and can result in decreased cell adhesion, changes in cell morphology and increased migration in vitro, and activation of ECM degrading enzymes including matrix-metalloproteinases (MMP).
Peritoneal Fluid and Anatomy
The peritoneal cavity is normally empty except for a thin film of fluid that keeps surfaces moist. Peritoneal fluid arises primarily from plasma transudate and ovarian exudate. Other sources of peritoneal fluid include fallopian tubal fluid, retrograde menstruation, and macrophage secretions. The volume of peritoneal fluid is usually 5–20 mL, and varies widely depending on physiological or pathological conditions. Peritoneal fluid contains a variety of free-floating cells, including macrophages, natural killer (NK) cells, lymphocytes, eosinophils, mesothelial (peritoneal surface epithelial) cells, and mast cells, which are all involved in immunological surveillance. Intraperitoneal fluid flow is directed by gravity to its most dependent sites and then drawn via the paracolic gutters to the diaphragm by the generation of negative intra-abdominal pressure in the upper abdomen during respiration. There is preferential flow along the right paracolic gutter, liver capsule, and diaphragm. Therefore, a natural flow of peritoneal fluid exists within the abdominal cavity, providing a route for the transcoelomic dissemination of detached tumor cells.
As the epithelial surfaces of the female genital tract (i.e., ovaries, fallopian tubes and endometrium) share a common embryological lineage with the peritoneal epithelium, it has been suggested that transcoelomic metastasis from gynecological malignancies, such as fallopian tube and ovarian tumors, are not true metastases but a result of malignant transformation at multiple foci throughout the peritoneum, i.e., peritoneal metaplasia. If the metaplasia hypothesis is correct, then one might expect metastatic lesions to be randomly distributed throughout the peritoneum. Alternatively, if the theory of dissemination via peritoneal/ascitic fluid is true, then one might expect that detached tumor cells would, by virtue of gravity, be more frequently implanted in the floor of the pelvis, e.g., the pouch of Douglas (the space between the rectum and back wall of the uterus), followed by the organs in the paracolic gutters, and finally on the diaphragm, i.e., along the normal route of peritoneal fluid circulation. Studies have shown that a high incidence of metastatic implants for all cancers, including ovarian malignancies, within the peritoneal cavity is found on organs where peritoneal fluid resorption occurs (omentum and omental appendages). In addition, the colon, greater omentum, and pouch of Douglas are most often affected, with a reduced incidence of implants seen on the small bowel and its mesentery, which is free to move by peristalsis, compared to the ileocecal area (the junction between the ileum and cecum), which is fixed to the retroperitoneum. Hence, location and topography with regard to the flow of peritoneal/ascitic fluid appear to be key determinants in the process of transcoelomic dissemination for all cancers. As such, in the case of gynecological cancers, peritoneal metaplasia alone appears unable to fully account for the peritoneal distribution of carcinomatosis.
Ascites: A Metastatic Milieu
The development of transcoelomic metastasis is often associated with the formation of excess peritoneal fluid known as malignant ascites. It is hypothesized that, in addition to hypoalbuminemia (low plasma albumin levels) secondary to dietary deficiency, at least three other pathological events can cause ascites: (1) reduced lymphatic drainage from the peritoneal cavity caused by the obstruction of lymphatic vessels by tumor cells; (2) increased vascular permeability of the peritoneal cavity; and (3) tumor neo-angiogenesis. While lymphatic obstruction is a well-recognized cause of ascites, the fact that massive amounts of fluid can accumulate in patients despite relatively little tumor burden suggests the involvement of other nonobstructive factors. These include vascular endothelial growth factor (VEGF), a glycoprotein which induces angiogenesis and increased vascular permeability in response to hypoxia. Other immune modulators, vascular permeability factors, and MMPs secreted by both tumor cells and mesothelial cells also contribute significantly to ascites formation and stimulate tumor growth, invasion, and angiogenesis.
Many immune cells, such as macrophages, are present in peritoneal fluid, and accumulate in so-called milky spots within the omentum. These omental macrophages have been found to be cytotoxic against tumor cells ex vivo. Consequently, omental macrophages might play an important role in killing tumor cells, thereby preventing development of transcoelomic metastasis and local peritoneal recurrences. Paradoxically, however, in vivo studies have shown that cancer cells seeded intraperitoneally specifically infiltrate the milky spots in the early stage of peritoneal metastasis. These studies suggest that omental milky spots are insufficient to prevent tumor progression, and that intraperitoneal metastasis requires tumor cells to possess or acquire mechanisms for evasion of immunological surveillance.
Tumor-infiltrating and malignant ascites-derived lymphocytes, in particular gamma–delta T cells, from patients with metastatic ovarian and colorectal cancer, have also been shown to possess antitumor activity. Hence, it appears that metastatic tumor cells have also developed strategies to evade T cell-mediated cytotoxicity. Fas ligand (FasL) is a transmembrane protein belonging to the tumor necrosis factor superfamily that can trigger apoptotic cell death following binding to its receptor, Fas. Expression of FasL has been observed in renal, ovarian, colorectal, and head and neck tumors and may be responsible for the immune privilege of tumor cells by inducing apoptosis of antitumor immune effector cells within the tumor microenvironment – the “Fas counterattack.” Studies have also shown that tumor progression and metastasis is associated with increased expression of FasL. Other examples of immune evasion include the recruitment of regulatory T (Treg CD4+ CD25+) cells to suppress tumor-specific T cell immunity; the presence of high concentrations of soluble forms of the complement pathway inhibitors C1 inhibitor, factor H and FHL-I on isolated metastatic ovarian cancer cells in ascitic fluid; and the phenomenon of spheroid formation observed in breast, colorectal, and ovarian cancer where tumor cells clump together by upregulating cell-adhesion molecules, thus resulting in increased complement resistance due to insufficient penetration of antibodies and complement into the spheroids.
Although topography appears to be a key determinant in the final site of metastatic implantation within the peritoneum, the actual mechanisms behind tumor implantation remain unclear. However, there is evidence to suggest the involvement of a dynamic regulation of the tumor cell’s adhesiveness, and its interaction with the underlying peritoneal mesothelium.
Potential mechanisms for the attachment of tumor cells to the peritoneal mesothelium include binding to ECM proteins like collagen type I and IV, laminin, and fibronectin via tumor cell surface integrins, and to hyaluronan expressed on the surface of human peritoneal mesothelial cells via the CD44 tumor cell surface protein, of which there are ten alternative exon splice variants (v1–v10). Upregulation of certain CD44 variants have been associated with distant metastasis in breast, colorectal, and ovarian cancer. Recently, tumor antigen/marker CA125, a glycoprotein overexpressed on the cell surface and secreted by ovarian tumor cells in the majority of ovarian cancer patients, has been shown to bind to mesothelin, a glycosylphosphatidylinositol-linked cell surface molecule expressed by mesothelial cells. Upregulation of the cell adhesion molecule E-cadherin may also mediate adhesion of circulating tumor cells to metastatic sites. Adhesion onto the peritoneal surface may be followed by haptotatic migration in which coordinated anti- and pro-migratory signals mediated by ECM proteoglycans confers directionality to tumor cell motility, effectively laying the tracks until a “stop” signal is encountered. Once attached to the peritoneal surface, metastatic cells proliferate and invade into the subjacent epithelium. The MMP family of proteinases and the urokinase-type plasminogen activator (uPA) system appear to be major contributors to this process. Human peritoneal epithelial cells and their associated immune and stromal cells have been shown to release regulatory chemokines and cytokines, such as IL-1, IL-6, and IL-8, in response to serosal inflammation and injury induced by tumor implantation, which in turn facilitate tumor angiogenesis and ascites formation (via increased secretion of VEGF), and enhanced tumor migration, attachment, proliferation, and invasion.
Finally, just as extraperitoneal tumors can metastasize to the peritoneum, intraperitoneal tumors can also metastasize extraperitoneally. Apart from the rich intraperitoneal network of blood and lymphatic vessels which can be invaded by tumors, peritoneal fluid is also continually being returned to the systemic circulation via the subdiaphragmatic lymphatic network and thoracic duct into the left subclavian vein, thus providing a direct “metastatic expressway” for peritoneal metastases to gain access into the lymphatic and circulatory system.
Patients with transcoelomic metastasis often present with signs and symptoms of abdominal pain, abdominal distension secondary to an enlarging tumor or ascites, constipation or diarrhea, shortness of breath, fatigue, loss of appetite, and weight loss. A careful clinical history followed by thorough clinical examination is required to ascertain the likely source of the primary tumor. Investigations should include routine blood tests, including relevant tumor markers, followed by radiological investigations including ultrasound and computer tomographic (CT) scans of the chest, abdomen, and pelvis to confirm the likely source of tumor and disease stage. In all cases, particularly those in which there is no obvious source of primary tumor (i.e., carcinoma of unknown primary origin), a biopsy of an accessible lesion should be obtained for histopathological and immunohistochemical confirmation and diagnosis.
In the past, clinical situations involving transcoelomic metastasis were treated mainly with palliative intent. Increasingly, studies have shown that an aggressive approach to peritoneal surface malignancy involving peritoneal debulking (cytoreductive) procedures, combined with optimal perioperative or postoperative systemic or intraperitoneal chemotherapy in carefully selected patients can result in long-term survival. Clinical assessment parameters that need to be considered include the patient’s performance status, preoperative abdominal and pelvic CT scans to define the extent and operability of disease, including the presence of extraperitoneal metastases, and tumor histopathology. Key prognostic indicators following surgery include the completeness of peritoneal debulking surgery, the presence of intraperitoneal lymph node and visceral metastases, and tumor type. Of the various scoring systems used to assess the extent of peritoneal carcinomatosis, the most frequently quoted is the peritoneal cancer index (based on the intraoperatively observed distribution and size of intraperitoneal metastasis) and the completeness of cytoreduction score (based on the amount of residual disease following peritoneal debulking surgery), which have been found to correlate well with prognosis in colorectal, gastric, and ovarian cancer. A meta-analysis of studies comparing combined peritoneal debulking surgery and perioperative intraperitoneal chemotherapy with systemic chemotherapy alone for the treatment of peritoneal carcinomatosis from colorectal carcinoma has demonstrated improved survival in the combination therapy group. In patients with ovarian cancer and peritoneal metastasis, 2-year survival following radical resection of all macroscopic tumors is 80%, in contrast to less than 22% for the patients with residual lesions larger than 2 cm. Early aggressive treatment of minimal peritoneal surface dissemination appears to confer the most benefit.
In patients with inoperable tumors at presentation, primary systemic or intraperitoneal chemotherapy is recommended, following which reassessment for surgical intervention may be possible if a good treatment response is observed. Palliative measures in the management of malignant ascites include repeated paracentesis (drainage of ascites), which provides relief in up to 90% of patients, and permanent percutaneous drains. The creation of a peritoneo-venous shunt (which allows ascitic fluid to drain from the peritoneal cavity into the superior vena cava) prevents the need for repeated paracentesis. Promising experimental approaches in the treatment of transcoelomic metastasis include the use of intraoperative hyperthermic intraperitoneal chemotherapy, anti-angiogenic agents such as the MMP inhibitors and the VEGF antagonists, as well as immunotherapy approaches including antibody-targeted T-cell therapy and combinations of intraperitoneal immunotherapy and thermochemotherapy.