Peritoneal Carcinomatosis from Gastric Cancer

Federico Bozzetti and Wansik Yu

The magnitude of the problem

Peritoneal carcinomatosis (PC) from gastric cancer is a frequent event even in the early phases of the disease. It has been estimated that a percentage of patients ranging from 15%[1] to 50% or more [2] have a peritoneal disease at the surgical exploration, especially when there is serosal involvement by the tumor.([3],[4]) These rates can dramatically increase if exfoliated neoplastic cells or their markers are actively searched in the peritoneal lavage cytology as well as if one considers autopsy findings.

PC, once established, is associated with a poor survival with median values ranging from 1-1.6 months([5],,[6]) to 3.1 and 9 months([1],[7]) and  no survival at 5 years.([8],[9])

Peritoneal recurrence is associated with serosal invasion[10], lymph node involvement[11], infiltrative growth[12], large tumor[13], gross appearance[14] and scirrhous-type reaction.[10]

Peritoneal dissemination is the most common reason for failure after intensive chemotherapy[15] and systemic chemotherapy is largely ineffective against peritoneal carcinomatosis.

Diagnosis and staging

The preoperative diagnosis and staging of peritoneal disease is limited by the insensitivity of traditional imaging modalities such as computed tomography (CT), ultrasonography (US), and magnetic resonance imaging (MRI). Although there is increased accuracy in the assessment of extragastric disease with modern multiphase, multidetector spiral CT imaging, sensitivity of CT for detection of extragastric disease declines with the size of metastases. Current CT techniques cannot consistently identify low-volume macroscopic metastases that are 5 mm or less in size. However, CT scans are sensitive for the detection of omental metastases or indirect evidence of tumor such as the presence of ascites, mesenteric thickening, or matting of loops of bowel.

Lee et al.[16] reported that EUS was more sensitive than combined US and CT scan examinations in diagnosing ascites.

Staging laparoscopy has become an accepted part of the pretreatment evaluation of patients who are thought to have advanced gastric cancer. Laparoscopy allows for direct inspection of the peritoneal and visceral surfaces for detection of CT-occult small-volume metastases. Staging laparoscopy also allows for assessment of peritoneal cytology[17] and intraperitoneal evaluation with adjunctive diagnostic techniques such as laparoscopic ultrasound (LUS). The accuracy of laparoscopy for peritoneal metastasis is reported to be 92%.[18]

Two major unresolved issues remain regarding the timing and extent of laparoscopy. Laparoscopy can be performed as a separate staging procedure before definitive treatment planning or immediately before planned laparotomy. LUS and extended laparoscopy are techniques that may increase the diagnostic yield of laparoscopy. LUS involves examination of the stomach, perigastric region, and peritoneal cavity using a laparoscopic ultrasound probe, whereas extended laparoscopy involves a more detailed laparoscopic examination of the perigastric region that includes laparoscopic examination of the lesser sac and retrogastric space.

The majority of patients with peritoneal carcinomatosis are diagnosed and assessed at laparotomy. Among several systems for assessing the distribution of peritoneal surface disease, peritoneal cancer index (PCI) is commonly used. This is a clinical summary of both lesion size and distribution of peritoneal surface malignancy. The lesion size (LS) is used to assess the size of nodules. An LS-0 score means that no malignant deposits are visualized. An LS-1 score signifies that tumor nodules less than 0.5 cm in the greatest dimension are present. An LS-2 score signifies tumor nodules between 0.5 and 5.0 cm present. An LS-3 signifies tumor nodules greater than 5.0 cm in any dimension present. If there is a confluence of tumor, the lesion size is scored as 3. In order to assess the distribution of peritoneal surface disease, the abdominopelvic regions are utilized. For each of these 13 regions, an LS score is determined (Fig. 1). The summation of the LS in each of the 13 regions is the PCI for that patient. A maximal score is 39 (13 x 3). It is basically a recording of what the surgeon visualizes as he explores the abdomen. PCI can be easily applied to the nodular type of peritoneal seeding from gastric cancer. However, it is difficult to assess the diameter of peritoneal surface disease in diffuse infiltrating type of dissemination.

Clinical pathway

Patients who are found to have occult metastatic disease at laparoscopy or laparotomy are considered incurable and only palliative procedures are recommended for these patients.(Type 3)

In patients with peritoneal carcinomatosis from colon cancer, the prognosis of patients with low PCI scores is significantly better than that of patients with high PCI scores. The value of the PCI as a selection tool in this patient population is obvious. However, PCI as a selection tool in gastric cancer has not been evaluated. Kim et al. reported that survival benefit can be obtained by a surgical resection and early postoperative intraperitoneal chemotherapy only when the patients had no hepatic metastasis in patients with peritoneal dissemination of T3 gastric cancer. (Type 3)

The peritonectomy is indicated for the patients with peritoneal dissemination which is considered to be able to macroscopically completely resected. Patients with a limited number of metastatic nodules distributed in the peritoneal cavity, and those without evidence of the involvement of the liver or distant lymph nodes are eligible for peritonectomy. Peritonectomy is contraindicated for patients with many big nodules distributed on the mesentery of small bowel. In other words, PCI should be converted to 0 by cytoreduction, that means completeness of cytoreduction score 0 (no visible peritoneal carcinomatosis remains after cytoreduction). Invisible dissemination already spreads everywhere in the peritoneal cavity, even if the complete resection of peritoneal dissemination is achieved by aggressive surgical cytoreduction.(Type 3)

A major role for intraperitoneal chemotherapy is the prevention of subsequent peritoneal carcinomatosis. Intraperitoneal chemotherapy is an important treatment option for patients with T3 or T4 lesions as well as for patients who are undergoing peritonectomy.(Type 3)
The treatment strategy composed of peritonectomy, intraoperative and early postoperative intraperitoneal chemotherapy is expected to be a powerful therapy for patients with peritoneal dissemination.
Locoregional treatment consists in peritoneal perfusion (with or without hyperthermia) as a single procedure or following cytoreductive surgery. The two procedures are not comparable since the first has usually been used in the earliest stages and sometimes as an adjuvant treatment, the second one has been exclusively used in advanced stages.

Intraperitoneal chemotherapy (IT)

A meta-analysis of 11 RCTs (1980-2003) involving 1161 patients receiving IT (usually without hyperthermia and mainly with  mitomycin C, but also platinum or fluorouracil) after curative resection for locally advanced gastric cancer showed a benefit in the overall survival (pooled odds ratio 0.51, 95% CI 0.40.- 060).(Type 1 )

More recently two phase II studies were published but the only new report of a RCT is from Yu et al.They reported on 248 patients randomized to early postoperative intraperitoneal chemotherapy with mitomycin C and 5-fluorouracil for 5 days versus control. There was a benefit in the 5 and 10-year survival in the treated arm which was mainly confined to stage III.

A multivariate analysis confirmed that IT was a statistically significant prognostic factor for survival. Rate of postoperative complications and length of hospital stay, but not mortality, were significantly higher in the treated group.

Peritonectomy (P) and hyperthermic intraperitoneal chemotherapy (HIPEC)

To our knowledge there are three institutional studies on P-HIPEC in patients with gastric cancer, one retrospective (42 patients)( 23), one prospective (49 patients)and one comparative non-randomized (34 patients).(30) The median survival ranged from 8 months to 10-11 months(23,29) and the 5-year survival ranged from 6%(23,30) to 16% according to Glehen et al.(29) Surgical mortality was acceptable, namely 2 to 7.1%.(23,29,30) Interestingly, a multivariate analysis showed that completeness of cytoreduction (and hence initial stages of peritoneal dissemination versus advanced stages) and absence of ascites were favourable prognostic factors for survival. More precisely, in patients with CC-0 or CC-1 median survival was 21.3 months and 5-year survival was 29.4%(29), and Yonemura reported quite similar figures: a complete cytoreduction was associated with a median survival of 19.2 months and a 5-year survival of 27%(23). It is noteworthy, however, that a complete cytoreduction was possible in only 44-51% of patients operated on by Yonemura(23) and Glehen.(29)(Type 3)