Introduction

Gastric cancer is a disease that can be treated by surgery; however, the prognosis remains poor for cancers that cannot be treated by surgery or in cases of recurrence. Nearly 700,000 people die each year from gastric cancer; thus, it is the second most deadly cancer [1]. The frequency of gastric cancer is particularly high in Asia, Eastern Europe, and South America, and 50,000 people die annually from gastric cancer in Japan [2]. In clinical trials, groups treated with chemotherapy for advanced gastric cancer (AGC) were shown to live longer than untreated groups [3–5].

In recent years, the efficacy of paclitaxel (PTX) administered every 3 weeks for AGC has been observed in phase II clinical trials, but high rates of adverse events such as neutropenia and peripheral neuropathy have also been recognized [6, 7]. In this context, weekly PTX has been reported to be superior in terms of safety and equivalent in efficacy [8, 9]. Because adverse events with weekly PTX are minimal, weekly PTX has often been used as second- or third-line chemotherapy; several phase I and II clinical trials have also been conducted using PTX in combination with other drugs in front-line treatment [9–12]. Ninomiya et al. reported a 39% response rate (RR) in a phase II clinical trial that combined weekly PTX with continuous intravenous infusion (c.i.v.) of 5-fluorouracil (5-FU) in AGC treatment [13]. Nagata et al. reported a 41% RR in a phase II clinical trial of weekly PTX plus cisplatin (CDDP), in which the recommended doses were set (80 and 25 mg/m2, respectively) in a phase I trial [14, 15].

To evaluate the additional effects of 5-FU in combination with weekly PTX plus CDDP, we planned a phase I trial with combined weekly PTX, CDDP, and 5-FU for AGC. Intravenous bolus injection was selected as the method of administration for 5-FU in order to obtain the benefits of this treatment, which does not require hospitalization or a port system. As a result, recommended doses of PTX, CDDP, and 5-FU were set at 80, 25, and 600 mg/m2, respectively, and a phase II clinical trial was established [16].

Methods

Study design and statistical methods

This was a multicenter, open-label, single-arm, phase II study conducted at 4 institutions in Japan. The primary endpoint was the RR of weekly combination therapy, and secondary endpoints were progression-free survival (PFS), overall survival (OS), and adverse events.

On the basis of the results of previous clinical trials of 2-drug combination therapy including weekly PTX conducted in Japan [13, 15], and 3-drug combination therapy of DCF [docetaxel (DOC), CDDP, and 5-FU] reported by Ajani et al. [17], the present study was designed to reject an RR of less than 20% as a clinically meaningless RR and provide a statistical power of 80% for assessing an efficacy of 38% for the regimen with an alpha error of less than 5%. The required sample size was determined to be 44. PFS was measured from the date of registration to the date of documented progression or death from any cause. OS was defined as the time from registration until death from any cause. PFS and OS were updated 6 months after the registration of the final patient and the medians were calculated using the Kaplan–Meier method.

Patients

The patients enrolled in this study had histologically confirmed non-resectable (metastatic) or recurrent gastric cancers. Other criteria for inclusion were as follows: Eastern Cooperative Oncology Group (ECOG) performance status (PS) 0–2; measurable lesion(s); no history of treatment with taxanes, platinum-based compounds, or intravenous 5-FU (1 regimen of pre-treatment with oral 5-FU agents was allowed, and adjuvant chemotherapy of 6 months or more after completion was not considered prior treatment); adequate bone marrow, liver, and renal functions, as defined by white blood cell count between 4000 and 12,000/mm3, absolute neutrophil count >2000/mm3, platelet count >100,000/mm3, hemoglobin >8.0 g/dl, aspartate aminotransferase (AST) and alanine aminotransferase (ALT) <2 times institutional upper limits (<3 with liver metastasis), total bilirubin <1.5 mg/dl, creatinine <1.5 mg/dl; expected survival period of 3 months or more; no active cancer in other organs; and aged at least 20 years. Patients with severe organ dysfunction, history of hypersensitivity, inflammatory disease, massive ascites, and/or pleural effusion or brain metastasis were excluded from the study. The present study was conducted in accordance with the principles of the Declaration of Helsinki and Good Clinical Practice guidelines. The study protocol was reviewed and approved by the Institutional Review Boards of each institution, and all patients gave their written informed consent to participate in the study.

Treatment

Paclitaxel (PTX) at 80 mg/m2 with 200 ml of 0.9% sodium chloride solution was administered by infusion over a period of 1 h followed by a bolus infusion of 5-FU at 600 mg/m2 and a 2-h infusion of CDDP at 25 mg/m2 with 500 ml of 0.9% sodium chloride solution. To prevent hypersensitivity reactions, pre-medication (intravenous dexamethasone at 16 mg; peroral diphenhydramine at 50 mg, and intravenous ranitidine at 50 mg or famotidine at 20 mg) was given 30 min before PTX administration. Each treatment cycle consisted of the administration of these agents once a week for 3 weeks (days 1, 8, and 15), and this treatment regimen was repeated every 4 weeks.

The decision to administer the combination therapy was based on the physical examination of the patient and data collected within 24 h, and the standards were set as follows: leukocytes >3000/mm3 (days 8, 15 >1500/mm3); neutrophils >1500/mm3 (days 8, 15 >1000/mm3); hemoglobin >8.0 g/dl; thrombocytes >75,000/mm3; AST and ALT <2 times institutional upper limits (<3 with liver metastasis); fever <38°C; PS 0–2; non-hematological toxicity Grade 0–2.

When these conditions were not met on day 1, the period of observation was extended by 1 week. Failure to meet these conditions after the 1-week extension or on day 8 or 15 meant that treatment was skipped on that day. Treatment doses in the next cycles were reduced for patients who developed Grade 4 leukopenia/neutropenia, febrile neutropenia or Grade 3 or higher non-hematological toxicity, and these dose regimens included Level-1 (PTX, CDDP, and 5-FU; 70, 20, and 500 mg/m2, respectively) and Level-2 (60, 15, and 400 mg/m2, respectively). Granulocyte colony-stimulating factor (G-CSF) was not administered prophylactically, but was permitted to be administered when Grade 4 leukopenia/neutropenia or febrile neutropenia developed.

The treatment protocol included 6 treatment cycles; however, treatment could be stopped when a complete response (CR) or partial response (PR) was confirmed. In addition, the treatment could be extended to 7 cycles or more on the basis of the judgment of the attending physician. However, administration was stopped under the following conditions: progressive disease (PD); decline in overall condition; treatment skipped twice in 1 cycle; withdrawal of consent or request for discontinuation; patient deemed unsuitable after registration or the attending physician deemed further treatment to be inappropriate. No restrictions were placed on follow-up treatments after the completion of this regimen.

Evaluation of response and toxicity

Pre-treatment evaluation included determination of baseline medical history, the physical condition of the patients, laboratory studies (complete blood cell counts, electrolytes, liver and renal function tests, and urinalysis), chest X-ray, abdominal computed tomography (CT), and electrocardiogram. CT or other imaging tests were also performed to examine the measurable lesion; these tests were repeated at the end of every treatment cycle or every 2 months in patients who stopped treatment before progression. The confirmed responses were classified by central review according to the Response Criteria in Solid Tumors version 1.0.

Toxicity was monitored and graded according to the Common Terminology Criteria for Adverse Events version 3.0. Patients underwent weekly laboratory tests and non-hematological toxicity tests during treatment.

Results

Patient characteristics

Between March 2006 and May 2009, a total of 46 patients with AGC were enrolled in the study. The characteristics of the 46 patients are summarized in Table 1. Median age was 71 years (range 48–85 years); 36 patients (78%) were male, 15 patients (33%) had non-resectable cancer, and all patients had a good PS (0–1). Common metastatic sites included the lymph nodes (70%) and liver (24%). Eighteen patients had a history of pre-treatment with oral 5-FU agents [17 patients received S-1 (15 as adjuvant therapies and 2 as first-line therapies) and 1 received doxifluridine as an adjuvant therapy]. One patient withdrew consent before the start of treatment; therefore, only 45 patients were assessed for efficacy and toxicity.

Table 1 Patient characteristics
Full size table

Treatment delivery

A total of 237 cycles of weekly PTX, CDDP, and 5-FU were administered, with a median of 4.5 cycles per patient (range 1–24). Sixteen patients (36%) completed 6 treatment cycles, and the doses were reduced to Level-1 and Level-2 in 12 and 9 patients, respectively. Median relative dose intensity for PTX, CDDP, and 5-FU was 67%. One patient continued treatment past the 24th cycle. Treatment was discontinued for the other 44 patients for the following reasons: confirmation of PD in 25 patients, skipping the treatment more than once in a single cycle owing to adverse events in 11 patients, decline in overall condition in 5 patients, completion of the planned 6 cycles of treatment in 2 patients, and upon request from the patient in one patient. Of these 44 patients, 30 received further lines of chemotherapeutic treatment and 2 received radiotherapy.

Efficacy

The response rate (CR + PR) was 56% [95% confidence interval (CI) 41–70%], with 2 patients showing a CR (Table 2). The RR among patients with lymph node metastasis was 59% (19/32), and that among patients with liver metastasis was 64% (7/11). The RR among patients with no prior treatment was 48%, and that among patients with a history of pre-treatment with oral 5-FU agents was 67%. The disease control rate [CR + PR + stable disease (SD)] was 82% (95% CI 71–93%). The median follow-up time was 332 days, during which the median PFS was 160 days (95% CI 145–214 days, event = 37) and the median OS was 552 days (95% CI 350–843 days, event = 22) (Figs. 1, 2).

Table 2 Tumor response
Full size table
Fig. 1
figure 1

Kaplan–Meier analysis of progression-free survival (PFS) for all 45 patients treated with paclitaxel, cisplatin, and 5-fluorouracil

Full size image
Fig. 2
figure 2

Kaplan–Meier analysis of overall survival (OS) for all 45 patients treated with paclitaxel, cisplatin, and 5-fluorouracil

Full size image

Toxicity

Toxicities observed during treatment are listed in Table 3. Grade 3–4 major hematological adverse events observed in the patients were neutropenia (76%), leukopenia (38%), anemia (27%), and febrile neutropenia (22%). Seven of 10 patients who developed febrile neutropenia were 75 years old or more, and all patients recovered soon with the administration of G-CSF and antibiotics. Eight of these 10 patients could continue the treatment after recovery, with dose reduction, while two stopped treatment because of PD. Grade 3 major non-hematological adverse events observed were exacerbation of PS (16%) and anorexia (13%). Grade 4 non-hematological adverse events and treatment-related deaths were not observed.

Table 3 Toxicity (n = 45)
Full size table

Discussion

Although the median survival of AGC when left untreated is 3–4 months, the survival has been shown to improve with adequate chemotherapy. Whereas 5-FU had been and remains the core of chemotherapy for AGC, the advent of CDDP led to a new era of combination therapies such as CF (CDDP and 5-FU) and ECF (epirubicin, CDDP, and 5-FU) [18, 19]. Recently, the V325 study reported the superiority of DCF over CF in the United States, and the REAL2 trial reported the non-inferiority of EOX (epirubicin, oxaliplatin, and capecitabine) compared with ECF in European countries [20, 21].

In Japan, the non-inferiority of S-1 compared with 5-FU alone was shown in the JCOG9912 trial, thereby validating the introduction of S-1 in AGC treatment [22]. Furthermore, S-1 was also shown to be useful in postoperative adjuvant chemotherapy for stage II–III gastric cancer [23]. In the meantime, the superiority of S-1 plus CDDP over S-1 alone was shown in the SPIRITS trial, demonstrating again the effect of adding CDDP to the core treatment [24]. Thus, the S-1/CDDP combination is the current standard in Japan as first-line treatment for AGC. However, whether this combination is valid for a patient who was treated with adjuvant S-1 and was subsequently found to have recurrence remains unclear. According to Shitara et al., the RR of S-1-based secondary treatment in patients with a history of S-1 therapy is 0%; while this is based on a retrospective analysis, opinions regarding the re-introduction of S-1 are generally negative [25]. The establishment of a new regimen that is active against S-1-refractory recurrences and could also be used for AGC patients who are not capable of having oral food intake is warranted.

The addition of another drug, as in the recent trend in Western countries, could be a solution. Indeed, DCF was superior to CF in every endpoint, including PFS (5.6 months), OS (9.2 months), and RR (37%). However, DCF was shown to be problematic because of adverse events, with Grade 3–4 neutropenia observed in 82%, febrile neutropenia in 29%, Grade 3–4 stomatitis in 21% and diarrhea in 19% [20]. In addition, the effect of docetaxel in pretreated patients had not been documented. We therefore turned to PTX, which had been explored more extensively in the second-line setting. The RR of weekly PTX reached 20% when administered as a single agent, but increased to about 40% when it was combined with either 5-FU or CDDP [8, 9, 13, 15]. Recently, 2 studies reported on the combination of weekly PTX, CDDP, and 5-FU (c.i.v.). The RRs in these studies were 48% and 50% [26, 27], and were close to the RR achieved in the present study. In addition, the incidences of Grade 3–4 neutropenia in these weekly regimens (3 and 21%) were lower than the incidence for the same three-drug combination given every 3 weeks (54%) [28].

In the present trial, we combined a 5-FU bolus with weekly PTX plus CDDP and observed an additional effect, achieving an RR of 56%. Rather surprisingly, the combination was active in patients who had been pretreated with oral 5-FU agents, including S-1. 5-FU is known for having two modes of action; one of them is time-dependent and is exerted mainly by continuous infusion or oral administration, whereas the other is concentration-dependent and is exerted by bolus administration. There is a possibility that cancer that had become refractory to oral 5-FU agents remained sensitive to the concentration-dependent action of 5-FU.

Despite the use of weekly administration in the present study, hematological toxicities were frequent and comparable to those of the V325 study. This was partly because of the way 5-FU was administered and the background of the patients. While the bolus 5-FU inhibited additional gastrointestinal toxicities, it led to enhanced bone-marrow suppression. In addition, the median age of the patients enrolled in the present trial was higher, at 71 years, when compared with the ages of the patients in the other 3-drug trials (median age 51–62 years), and most of our patients who developed febrile neutropenia were over 75 years of age. Nevertheless, these toxicities were managed without difficulties through dose modifications and alterations in the treatment schedule.

On the other hand, our treatment regimen was associated with a low incidence of subjective adverse events despite being a triplet regimen. While median PFS was 160 days, which is no better than that reported for other existing clinical trials, the median OS was extremely good, at 552 days. As well as the fact that our study population had favorable characteristics, with a low rate of peritoneal metastasis and a high rate of recurrent cases with a small tumor volume, the low incidence of subjective adverse events may have contributed to enhance the OS time. Few patients experienced a decline in general condition during the present treatment and, consequently, further lines of treatment were delivered to several of the patients after termination of the protocol treatment.

In conclusion, the results of the present phase II trial show that the combination of weekly PTX plus CDDP and bolus 5-FU is effective against AGC. While adequate attention to neutropenia and febrile neutropenia is required, especially for elderly patients, we consider that the adverse events observed with this combination therapy are within the permissible range. We believe that this regimen is promising as a first-choice therapy for AGC, especially in patients who have difficulties with oral ingestion or those who have received prior treatment with oral 5-FU. We intend to further investigate the usefulness of this combination in a randomized controlled trial.