ORLANDO, FL—Positive peritoneal cytology in itself is a poor predictor of survival in patients with gastric cancer. But a study from Memorial Sloan-Kettering Cancer Center has found that patients with positive cytology and no evidence of metastasis have significantly longer disease-specific survival than those with positive cytology and metastatic disease.
Diffuse tumors and performance status were also associated with shorter disease-specific survival, said the study's first author, James J. Mezhir, MD, a surgical oncology fellow, reporting the results here at the Gastrointestinal Cancers Symposium.
“Although long-term survival in patients with positive cytology is extremely uncommon, those who convert to negative cytology after chemotherapy exhibit significant improvement in disease-specific survival,” Dr. Mezhir said, in an oral abstract session on cancers of the esophagus and stomach.
Cytology-positive patients with no metastases had a median disease-specific survival of 1.3 years (range of 0.3 to 5.8 years), compared with 0.8 years (range of 0.01 to 3.7 years) for cytology-positive patients with metastatic disease. The hazard ratios were 1.0 and 1.9, respectively.
The study was conducted to identify variables at known initial presentation that predict disease-specific survival from the time of initial staging laparoscopy in patients with positive cytology.
“We wanted to examine the natural history of patients with positive cytology as the only evidence of metastatic disease, and also to examine a subgroup of patients who had repeat cytology after treatment with systemic chemotherapy,” he explained.
Study data were taken from a prospectively maintained gastric cancer data base at the cancer center.
From 1993 to 2009, 1,241 patients with gastric cancer underwent diagnostic laparoscopy with peritoneal washings. Of those, 291 (23%) had positive cytology.
A total of 198 of these patients had visible peritoneal or visceral metastasis discovered at laparoscopy. The remaining 93 had positive cytology but no visible visceral or peritoneal metastasis.
Median disease-specific survival was one year for the entire cohort, and all but one patient died of disease by 58 months.
Seventy two patients (25% of the cohort with positive cytology) were selected for resection of their primary tumor. Among those, the majority, 55 patients (76%), had T3 disease. Another 11 patients (15%) had T1/2 disease, and six (8%) had T4 disease.
A total of 86% of that cohort had nodal disease. “Resected patients were more often without metastases—72% versus 19%,” he said.
Of 261 patients treated initially with chemotherapy, 48 (18%) had a repeat laparoscopy for re-evaluation. Of those, 21 had persistent positive cytology (median disease-specific survival 1.4 years) and 27converted to negative cytology (median disease-specific survival of 2.5 years).
“Clearly, patients who convert to negative cytology after systemic chemotherapy have a significant increase in their disease-specific survival,” Dr. Mezhir said.
Tumor Location, Performance Status
Patients with tumors in the antrum had a median disease-specific survival of 1.2 years, compared with 0.6 years for patients with diffuse disease.
And a good performance status was associated with a median disease-specific survival of 1.07 years, vs 0.47 years for those with poor performance status.
The Discussant for this report, Kevin C. Conlon, MD, Professor of Surgery at Trinity College at the University of Dublin, called it a provocative study that provides ammunition to develop further trials.
“The real question Dr. Mezhir and colleagues have raised is, what is the optimal therapy for patients with localized disease and no evidence of metastasis.”
He said many surgeons have the attitude that a smaller tumor can be resected with curative intent, but the literature shows that surgery alone is not curative for these patients. “And there have been a multitude of attempts at developing neoadjuvant approaches, cytoreductive therapy, or intraperitoneal therapy—most of these trials from Asia and Japan,” Dr. Conlon said.
Dr. Conlon complimented Dr. Mezhir and colleagues for using a prospective data base, because it allows researchers to interrogate patient populations, as they have done.
And he also complimented the authors for finding that the 93 study patients who had no visible disease but did have positive cytology had a significantly longer survival than patients with visible metastasis.
“What this very important paper has confirmed is that peritoneal cytology is an important prognostic variable and should be part of our staging algorithm,” Dr. Conlon said. “It also showed us that patients who respond to chemotherapy have improved survival.”
A question still unanswered is what the appropriate therapy should be for these patients, he said, and does surgical resection impact survival.
“We need prospective randomized trials with survival as the primary endpoint. Unfortunately, using the RECIST criteria is problematic because the vast majority of patients do not show a response to the primary tumor.”