Patients had histologically confirmed metastatic gastric adenocarcinoma. Eligibility criteria included at least one measurable lesion; Eastern Cooperative Oncology Group performance status ≤2; age >18 yr, life expectancy ≥3 months, no other malignancy; and adequate hematological, renal, and hepatic function (white blood cell count ≥3,000/µL, granulocytes ≥1,500/µL, platelets ≥100,000/µL, total bilirubin <1.3 mg/dL, and creatinine clearance >60 mL/min). No prior palliative chemotherapy was permitted in patients with advanced disease. Adjuvant chemotherapy was allowed if >12 months had elapsed between the end of therapy and registration.

Exclusion criteria were as follows: preexisting neurotoxicity greater than or equal to grade 2 of the National Cancer Institute Common Toxicity Criteria (NCI-CTC, version 2.0), concurrent cancer, central nervous system metastasis, active infection, other serious underlying medical conditions that would impair the ability of the patient to receive the planned treatment, or inadequate calorie and fluid intake. Participants gave written informed consent before they entered the study, which was approved by the Chonnam National University Hospital ethical committee.

Paclitaxel (175 mg/m²) and cisplatin (75 mg/m²) were given as a 1-hr intravenous infusion on day 1, followed by 5-FU (750 mg/m²) as a 24-hr continuous infusion for 5 days. Cycles were repeated every 3 weeks. Thirty minutes prior to the paclitaxel infusion, each patient received 20 mg dexamethasone, 50 mg ranitidine, and 5 mg chlorpheniramine maleate intravenously to prevent hypersensitivity reactions. After prehydration with at least 1 L normal saline, the calculated dose of paclitaxel, diluted in 300 mL normal saline, was infused over 1 hr. The calculated dose of cisplatin, diluted in 500 mL normal saline, was then administered over 1 hr, followed by posthydration with 3 L normal saline over 24 hr. Soon after finishing the cisplatin infusion, 5-FU was infused continuously for 5 days. Ondansetron (8 mg, i.v) was routinely given. Patients received further cycles of chemotherapy only when the absolute neutrophil count was ≥1,000/µL and platelets were ≥150,000/µL. Toxicity was graded according to NCI-CTC version 2.0. Dose modifications were determined by hematological or non-hematological toxicities. The cisplatin dose was reduced to 75% of the original dose in subsequent cycles if one of the following occurred: grade 3 neutropenia with infection, grade 4 neutropenia, grade 3 thrombocytopenia with bleeding that required platelet transfusion and greater than grade 3 sensory neurotoxicity, or grade 2 or greater nephrotoxicity. In case of subsequent toxicities after these dose reductions, the paclitaxel dose was reduced to 75% of the original dose. In cases of asthenia above grade 3, treatment was postponed for 1 week and restarted when the patient had recovered to below grade 2. If the toxicity did not improve to below grade 1 after 4 weeks, the patients was removed from the study. Also, the paclitaxel dose was reduced to 75% of the original dose if grade 2 sensory neurotoxicity was sustained subsequent 2 cycles despite the medication such as amitryptiline. For patients with grade 3 diarrhea that lasted for more than 7 days despite the administration of loperamide or mucositis grade 3 lasting for more than 5 days or grade 4 mucositis, a 25% reduction in the daily dose of 5-FU was required. To avoid the delay of chemotherapy, G-CSF was given if the absolute neutrophil count was lower than 1,000/µL seven to ten days after chemotherapy. Patients showing disease regression or stable disease received additional cycles of chemotherapy, up to a maximum of nine cycles, unless they experienced disease progression or unacceptable toxicity, or if the patients refused or the physician decided against further treatment.

The antitumor activity was evaluated according to the World Health Organization (WHO) criteria. Computed tomography (CT) scans of measurable lesions were carried out within 4 weeks before the start of the treatment and were repeated after every three cycles. Complete response (CR) was defined as the complete disappearance of all evaluable disease, persisting for >4 weeks. Partial response (PR) was defined as a ≥50% reduction in the sum of the products of the largest perpendicular diameters in all measurable lesions for ≥4 weeks, without evidence of progression or the development of new lesions. Progressive disease (PD) was defined as an increase in a previous lesion by >25%, or the development of any new lesion, and stable disease (SD) as any change in a previous lesion that did not fit into either the PR or PD categories. The patients were consulted in assessing the response if they had early disease progression or had received at least two cycles of treatment.

The primary endpoint was ORR, and secondary endpoints were PFS, OS and toxicity. This study was conducted using a Simon two-stage design. The optimal number of patients was determined assuming a target level of interest, p₁=0.6, and a lower activity level, p₀=0.4. Initially we planned to enroll 34 patients in first stage. If 18 or more responses were observed, we planned to continue second stage for a total of 39 evaluable patients. This design provides for a probability of ≤0.05 of accepting drugs worse than p0, and a probability of ≤0.20 of rejecting drugs better than p1. If we assume the dropout rate to be 15%, the total number of enrolled patients will be 45. All enrolled patients were included in the intent-to-treat analysis. The Kaplan-Meier method was used in all survival analysis. The Statistics Package for the Social Sciences (SPSS) program for Windows (Version 13.0) was applied for analysis.

Forty-five patients were initially enrolled in this study. Of these, four patients were dropped out; one had another primary tumor (esophageal cancer), two were lost to follow-up after one cycle of chemotherapy, and one did not want further chemotherapy after one cycle. The median age of patients was 56 yr, with a range of 21 to 75 yr (Table 1). Thirty patients were male (68%) and 14 were female (32%); most patients were in good general condition with a performance status of 0 or 1 (82%). All patients had histologically confirmed adenocarcinoma. Twenty-seven patients (61%) had recurrent disease after operation with or without adjuvant chemotherapy and seventeen patients (39%) had no prior treatment in metastatic gastric cancer. The median follow-up duration was 12.8 months (95% CI, 6.2-24.8)

The median number of treatment cycles was six, with a range from three to nine cycles. The number of patients receiving more than six cycles was 33 (80%). The median duration of treatment for all patients was 147 days, with a range of 46 to 245 days. All patients received more than three cycles of chemotherapy, and forty one patients were able to evaluate their response. The total number of cycles delivered was 269. The number of patients who were received at least 1 dose reduction and at least 1 cycle delay were 13 (29.5%) and 7 (15.9%), respectively. The target doses for paclitaxel, cisplatin, and 5-FU were 58.3, 25, and 1,250 mg/m²/week, respectively, and the achieved doses for paclitaxel, cisplatin, and 5-FU were 52.3 (89.7%), 22.3 (89.2%), and 1,131.2 mg/m²/week (90.5%), respectively.

Initially, 34 patients were treated as scheduled. Of these patients, one patient showed CR, 18 patients showed PR, 12 patients showed SD and 3 patients showed PD. Therefore, we advanced this study up to 45 patients. The best responses are shown in Table 2. Two patients achieved CR (4.9%; 95% CI, 1-16) and 19 patients PR (46.3%; 95% CI, 31-63). The ORR was 51.2% (95% CI, 35-67) with a median response duration of 23 weeks. The responses of the 26 patients who had recurrent disease were 2 CR, 11 PR, 9 SD, and 4 PD, and the responses who had no prior treatment in metastatic disease were 8 PR, 6 SD, and 1 PD. No statistically significant difference was observed between the two groups. Among 12 patients who had carcinomatosis peritonei with ascites, four showed marked reduction of ascites after chemotherapy, and three did not need further paracentesis (SD). Two patients were treated with palliative surgery after the sixth and ninth cycles, and they had no evidence of disease until data analysis.

Of the 44 patients, four (9.1%) remained alive at a median follow-up of 12.8 months. The median progression-free survival was 6.9 months (95% CI, 5.86-7.94). The median OS was 12.7 months (95% CI, 9.9-15.5) and the 1-yr OS rate was 62.5% (Fig. 1).

Chemotherapy-related toxicity is summarized in Table 3. The main hematological toxicity was neutropenia. Grade 3 or 4 neutropenia was observed in 67 cycles (25%) and 54% of patients. Greater than grade 3 thrombocytopenia developed in only two patients (5%). Eighteen patients (41%) received growth factor before starting the next cycle of chemotherapy to prevent delayed treatment. Febrile neutropenia developed in three patients (7.3%) and all of them recovered after treatment; two patients recovered after a short course of antibiotic treatment and one patient needed intensive care treatment. Grade 1 or 2 peripheral sensory neuropathy was observed in 32 patients (72%), and greater than grade 3 peripheral sensory neuropathy occurred in two patients (5%). The major non-hematological toxicity was asthenia, which spontaneously resolved after 1 or 2 weeks of rest. No treatment-related mortality occurred.