Comparison of efficacy of different route of administration of chemotherapy on unresectable, advanced gastric cancer

The study was approved by Ethics committee of Jinling Hospital and informed consent was obtained from all patients.

The study enrolled patients with unresectable, advanced gastric cancer who were hospitalized from January 2007 to December 2009. Cancers meeting the following criteria were diagnosed as unresectable gastric cancer:1) strongly suspicious for stage III or IV metastasis of lymph nodes by enhanced CT examination; 2) tumor infiltration and encompassment of major blood vessels (for example, hepatic artery, celiac artery, and portal vein); 3) distant metastasis (for example, liver metastasis). The inclusion criteria were:age35 to75 years; Eastern Cooperative Oncology Group (ECOG) score of 0 to 2; no previous history of curative or palliative surgery, radiotherapy or chemotherapy; no serious cardiovascular, liver, or kidney disease; and acceptance of chemotherapy and interventional chemotherapy. The exclusion criteria were: pregnancy or breast-feeding; presence of peritoneal seeding or distant metastasis except in the liver; presence of other malignant tumors; history of curative or palliative surgery, radio therapy, or chemotherapy; history of upper gastrointestinal bleeding, perforation, obstructive jaundice, severe infections, or other serious complications; history of any serious or uncontrollable venous diseases; or sensitivity to any of the chemotherapy drugs.

In total, 85 patients (62 men, 23 women, 61 ± 13 years, range 35 to 75 years) were enrolled in the study. The diagnosis of gastric cancer was confirmed by histopathological examination of an endoscopic biopsy sample. The clinical staging of all cases were confirmed by CT and endoscopic ultrasonography before treatment. The relationship between the tumor and major blood vessels or organs in its proximity, and the extent of lymph-node metastasis were assessed by multi-slice spiral CT and endoscopic ultrasonography.

The selected cases were randomly divided into two groups to receive the appropriate preoperative chemotherapy regimens: 48 patients received the FLEEOX regimen, which combined arterial with venous administration, and 37 patients received the XELOX regimen, which was given by venous administration only.

The FLEEOX regimen consisted of a slow intravenous infusion of 5-fluorouracil(5-FU) 370 mg/m²over 5 days and an intravenous infusion of calcium folinate 200 mg/m²over 5 days. Following this, oxaliplatin 120 mg/m², epirubicin 30 mg/m², and etoposide 70 mg/m² were injected into the tumor site on days 6 and 20. Patients received one or two cycles of this regimen; for those patients receiving two cycles, the second cycle was administered after an interval of 5 weeks. The arterial administration was performed using the Seldinger method. After intubation of the celiac artery, a catheter was inserted into the blood supply of the tumor depending on the tumor site; for example, the catheter was inserted into the left gastric artery for cancer of the upper and central stomach, and through the hepatic and gastroduodenal arteries into the right gastroepiploic artery for cancer of the lower part of the stomach. First, half the drug volume was injected into the artery supplying the tumor, then the remaining drug was injected into the celiac artery, except for patients with accompanying liver metastases, for whom the second half of the drug was injected into the metastatic focus in the liver. Lipiodol embolization was also performed for the latter patients.

The XELOX regimen comprised intravenous infusion of oxaliplatin 130 mg/m²over 2 hours on day 1, followed by capecitabine 1000 mg/ m² orally twice daily for 2 weeks. This cycle was repeated once every 3 weeks, and patients were given two to four cycles.

The efficacy was evaluated by CT after three cycles of intervention treatment in the FLEEOX (combined arterial and venous administration), and three cycles in the XELOX (intravenous administration) group. Patients whose tumors were deemed resectable would have their chemotherapy stopped and surgery performed. Patients whose tumors were considered unresectable would continue chemotherapy for two cycles (FLEEOX group) or four cycles (XELOX group), after which the efficacy was again evaluated by CT. At this point, any patients with resectable tumors would undergo surgery. The study would be stopped for any patients whose tumors were not resectable.

The efficacy was evaluated by two radiologists with using the Response Evaluation Criteria In Solid Tumors (RECIST) criteria[12]. Tumors were evaluated as follows: complete disappearance of the tumor was considered to be complete response (CR); a decrease of more than 30% in tumor size to be partial response (PR), and an increase of more than 20% in tumor size as progressive disease (PD), neither sufficient shrinkage to qualify for partial response nor sufficient increase to qualify for progressive disease was considered stable disease (SD). Clinical RR was calculated as: (CR + PR)÷measurable number of cases) × 100%. Surgery (mainly D2 gastrectomy) was carried out within 2 weeks of the end of preoperative chemotherapy for patients with resectable gastric cancer. D3 gastrectomy was performed for patients with tumors andN3 lymph-node metastasis.

Patients were closely monitored for liver and kidney function, bone-marrow hematopoiesis, gastrointestinal reactions, and related adverse events during the treatment. Toxic reactions were evaluated using the National Cancer Institute Common Toxicity Criteria (version 3.0) and compared between two groups.

After chemotherapy, patients attended a clinical follow-up review every 3 months, which included physical examination, routine blood investigations, assessment of liver and kidney function, tumor markers, abdominal CT, and chest X-ray. Gastroscopy was performed once every year.

The primary objective indicators were clinical objective RR (CR + PR), 1-year and 2-year disease-free survival (DFS) rates, and median survival time. The survival rate was calculated by the Kaplan-Meier method. Secondary indicators were R0 resection rate, characteristics of the resected specimen, and toxic reactions. Overall survival (OS) time was recorded from first chemotherapy to death or the last follow-up visit. Comparison between the two groups was performed using the χ² test. Data were analyzed using the SPSS statistical software (version 17.0; SPSS Inc., Chicago, IL, USA). P < 0.05 was considered significant.

Of the forty-eight patients in the FLEEOX group, six patients were rated as having CR, thirty-five as PR, five as SD, and two as PD; the clinical RR was 85.4%. For the thirty-seven patients in the XELOX group, the figures were one, twenty, twelve and four, respectively, and the clinical RR was 59.5%. There was a significant difference in RR (P < 0.02) between the two groups (Table2).

In the FLEEOX group, the tumors were evaluated as resectable in nine patients after the second treatment cycle and in thirty-two patients after the third cycle. Of the remaining seven patients, six were evaluated as invalid or having disease progression, and one patient refused surgery. Of the forty-one patients approved for surgery, thirty-eight underwent D2 gastrectomy and three underwent D3 gastrectomy. The gastrectomy was a distal subtotal gastrectomy in twenty patients and a total gastrectomy in twenty-one patients. Thirty-six of the patients underwent R0 resection (thus theR0 resection rate was 75%; 36/48), while four patients underwent R1 resection and one underwent R2 resection.

The thirty-seven patients in the XELOX group received two (n = 5), three (n = 11) or four (n = 21) cycles of chemotherapy. Sixteen patients were evaluated as invalid or having disease progression, and thus could not undergo surgery. The remaining twenty-one patients who were approved for surgery all underwent D2 gastrectomy, with twelve operations being distal subtotal gastrectomy and nine being total gastrectomy. Seventeen patients underwent R0 resection (R0 resection rate of 45.9% (17/37)), while one patient underwent R1 resection and three underwent R2 resection.

There was a significant difference between the two groups in the R0 resection rate (P < 0.001)(Table3).

Upon intra-operative and post-operative observation of the primary tumors and metastatic lymph nodes, atrophic changes were found in 27 primary tumors in FLEEOX group, which occurred as caseous necrosis and scar-like changes in the corresponding lymph nodes. Necrosis of the gastric mucosa was visible in the peri-tumoral arterial administration area in 17 patients. Severe cases showed obvious scar-like changes (Figure1).

Atrophic changes were found in 13 primary tumors in the XELOX group, which produced caseous necrosis on the corresponding lymph nodes, but the gastric mucosa in the peri-tumoral area was normal in all patients.

Between the two groups, there were 76 patients who received D2 gastrectomy. There were no deaths due to the surgery. There were seven (16.3%) instances of post-operative complications in the FLEEOX group (two cases of intestinal obstruction and two cases of abdominal infection, and one case each of anastomotic leakage, pneumonia, and wound infection), and six (18.1%) in the XELOX group (two cases of abdominal infection, and one case each of intestinal obstruction, anastomotic leakage, pneumonia, and wound infection). There was no significant difference in the incidence rate of postoperative complication of the patients in two groups (P >0.05), and all patients recovered with conservative treatment.

In total, there were 207 completed cycles of chemotherapy for the two groups. No treatment termination or death occurred as a result of toxic reactions. During the chemotherapy, there were different degrees of toxic and adverse reactions in the two groups. The reactions were mainly myelosuppression, liver dysfunction, and gastrointestinal reactions. With regard to bone-marrow suppression, the levels of hemoglobin, white blood cells, and platelets decreased in some patients within both group. The most common symptom was leukopenia (n = 30; 62.5%) in the FLEEOX group and anemia (n = 24; 64.9%) in the XELOX group. Alanine aminotransferase (ALT) and aspartate aminotransferase (AST), which are markers of liver injury, increased in nine (18.8%) and eleven patients (22.9%), respectively, in the FLEEOX group, and in nine (24.3%) and eight patients (21.6%) in the XELOX group. The commonest digestive-tract reactions in both groups were varying degrees of nausea and vomiting, with 38 patients (79.2%) in the FLEEOX group reporting nausea,22(57.9%) patients vomiting, and 19 patients (51.4%) in the XELOX group reporting vomiting,25(67.7%) patients nausea. Toxic neurological reaction occurred in 21 patients (56.8%) in the XELOX, but in only 11 patients (22.9%) in the FLEEOX group. For most patients, there were only minor changes in renal function, with no marked change in levels of creatinine or urea nitrogen before and after chemotherapy in either groups. There was no significant difference in the number of toxic reactions between the two groups (P >0.05) (Table4).

The survival time of the 85 patients was calculated. The median survival time was 25 and 9 months in the FLEEOX group and the XELOX group, respectively, which was significantly different (P < 0.001) (Figure2). The 1-year and 2-year DFS rates were 85.4% and 45.8%, respectively, in the FLEEOX group, and 35.2% and 8.3%, respectively, in the XELOX group. The 1-year and 2-year DFS rates were 83.3% (40/48) and 33.3% (16/48) respectively in the FLEEOX group, with the number of recurrences in R0 patients being two and eighteen in the first year and second years, respectively. For the XELOX group, the 1-year and 2-year DFS rates were 33.3% (16/48) and 5.4%(2/37)respectively, and the number of recurrences in R0 patients were three and twelve in the first and second years, respectively. There was a significant difference in DFS rates between the two groups (P < 0.05) (Figure3).