A phase II study of paclitaxel, weekly, 24-hour continuous infusion 5-fluorouracil, folinic acid and cisplatin in patients with advanced gastric cancer

To evaluate the toxicity and efficacy of combination chemotherapy with paclitaxel, cisplatin and 24 h continuous infusion of 5-FU/folinic acid in patients (pts) with unresectable, locally advanced or metastatic gastric adenocarcinoma. Forty-five chemotherapy-naive pts (28 male and 17 female) with a median age of 60 years (range 35–74) were enrolled. 5-FU 2 g/m2was given weekly over 24 h i.v. preceded by folinic acid 500 mg/m2as a 2 h infusion. Paclitaxel 175 mg/m2was administered as a 3 h-infusion on days 1 and 22 and cisplatin 50 mg/m2as 1 h infusion on days 8 and 29. Six weeks of therapy (days 1, 8, 15, 22, 29, 36) followed by 2 weeks rest were considered one cycle. A median of 3 cycles (range 1–4) were administered to 45 pts assessable for response, survival and toxicity. Five pts (11%) obtained a CR and 18 pts (40%) a PR (ORR 51%; 95% Cl: 35.8–66.3%). Responses were achieved in the liver, lymph nodes, lungs and at the site of the primary tumour. Nine pts (20%) had stable disease. Thirteen pts (29%) were considered to have failed treatment, 8 pts (18%) due to progressive disease and 5 pts (11%) who did not receive one complete cycle of therapy due to acute non-haematologic toxicity. The median progression-free and overall survival times were 9 months (range 1–36+) and 14 months (range 2–36+), respectively. Neutropenia WHO III°/IV° occurred in 7 pts (15%) with only 1 pt having grade IV. Additional non-haematologic WHO III°/IV° toxicities included nausea/vomiting in 5 (11%), alopecia in 22 (49%), and diarrhoea in 1 patient each (2%). Dose reductions or treatment delays were necessary in 8 pts (17%), mainly due to neutropenia. All pts were treated on an outpatient basis. The combination of paclitaxel, cisplatin and continuously infused 5-FU/folinic acid appears to be a highly active regimen for the treatment of pts with advanced gastric cancer. While the overall acceptable toxicity allows its use in the palliative setting, it may also be an attractive option to be tested for neoadjuvant or adjuvant treatment. © 2000 Cancer Research Campaign

Despite a declining incidence during the past 60 years, gastric cancer is still among the most common cancers in Europe (World Health Organization, 1998). Metastatic gastric cancer remains an incurable disease with a median survival time of only 4-8 months.
Randomized studies have demonstrated a survival benefit and impact on quality of life for patients with irresectable or metastatic gastric cancer treated with chemotherapy plus supportive care as compared to best supportive care alone (Pyrhönen et al, 1995;Glimelius et al, 1997). Thus, chemotherapy for advanced gastric cancer is now widely accepted in Europe and the US. A number of drugs have demonstrated activity such as 5-fluorouracil (5-FU), etoposide, doxorubicin, methotrexate and cisplatin. Combination regimens, usually based on 5-FU, have achieved response rates between 20 and 40% (Schipper and Wagener, 1996;Wils, 1996). Several studies have shown that the combination of 5-FU and cisplatin is active and well tolerated in patients with gastric cancer (Okada et al, 1991;Kim et al, 1993;Rougier et al, 1994;Wilke et al, 1996a).
Paclitaxel is an antimitotic agent with the unique cytotoxic mechanism of tubulin stabilization and polymerization, resulting in nonfunctional microtubules (Rowinsky et al, 1990). Paclitaxel exhibits an antitumour activity against various tumours including gastric cancer cell lines (Vanhoefer et al, 1995;Chang et al, 1996). Ajani et al obtained a 17% response rate using paclitaxel as a single agent in gastric cancer, with a tendency towards a higher response rate in patients receiving paclitaxel as continuous infusion over 24 hours. A 22% response rate was observed in pretreated patients with gastric cancer by Cascinu et al, with paclitaxel as single agent administered over 3 hours Cascinu et al, 1998). Studies investigating the combination of paclitaxel with 5-FU revealed response rates between 15 and 50%, associated with a rather low toxicity profile (Cascinu et al, 1997;Murad, 1999). We have reported the results for the combination of 5-FU given as a weekly 24-hour continuous infusion plus folinic acid for 6 consecutive weeks in combination with paclitaxel at 3 weekly intervals (Bokemeyer et al, 1997a). A response rate of 32% was observed in 22 chemonaive gastric cancer patients resulting in a progression-free and overall survival of 8 and 11 months, respectively. Toxicity was mild with neutropenia (14%), alopecia (45%) and nausea and vomiting (5%) being the most frequent WHO grade III/IV toxicities (Bokemeyer et al, 1997a).
A phase II study of paclitaxel, weekly, 24-hour continuous infusion 5-fluorouracil, folinic acid and cisplatin in patients with advanced gastric cancer Two recent studies examined the combination of paclitaxel, cisplatin and 5-FU in patients with locally advanced adenocarcinoma of the oesophagus in the neoadjuvant setting, as well as in patients with metastatic disease. A high response rate combined with an acceptable toxicity was reported (Belani et al, 1997;Ilson et al, 1998).
Based on the activity of 5-FU and cisplatin in gastric cancer and on the favourable results of the combination of 5-FU/folinic acid and paclitaxel, we here present a phase II study combining weekly 24 h continuous infusion 5-FU/folinic acid with cisplatin and paclitaxel in an alternating schedule in patients with gastric cancer.

Patients
Patients with histologically proven advanced or metastatic gastric cancer were included in this study. Inclusion criteria were as follows: unresectable, locally-advanced or metastatic disease; ECOG performance status 0-2; presence of measurable disease; age between 18 and 75 years; life expectancy > 3 months; no prior chemotherapy; adequate haematological, renal and hepatic function as defined by a granulocyte count ≥ 1.5 × 10 9 /l, thrombocytes ≥ 100 × 10 9 /1, serum creatinine ≤ 1.5 mg/dl, creatinine clearance ≥ 60 ml/min, bilirubin level < 2-fold and liver enzymes < 3-fold the upper normal limits and written informed consent. All patients had to be available for follow-up.
Patients were excluded from the study in case of one of the following: active bleeding, diffuse, non-measurable liver metastases; history of a secondary malignancy except for nonmelanomatous skin cancer or carcinoma in situ of the cervix; concurrent insufficiently treated disease such as heart, renal or hepatic failure or uncontrolled infection; prior chemotherapy; presence of a concurrent psychiatric disorder; pregnancy.
Prior to therapy, a clinical history and physical examination, a complete blood count and serum chemistry including liver and kidney function tests, a creatinine clearance as well as an electrocardiogram and an audiogram were obtained. All measurable tumour lesions were radiologically assessed either by CT-scan, chest-X-ray or abdominal ultrasound.
The study was approved by the Ethics Committee of the University of Tuebingen.

Treatment plan
Treatment was given once weekly for a total of 6 weeks followed by 2 weeks rest. This period was defined as one treatment cycle ( Figure 1). 5-Fluorouracil was administered weekly at a dose of 2000 mg/m 2 as 24 h-continuous infusion preceded by 500 mg/m 2 folic acid as a 2 h infusion. Paclitaxel was administered at a dose of 175 mg/m 2 on days 1 and 22 as a 3 h infusion. All patients received dexamethasone 20 mg, ranitidine 300 mg as well as diphenhydramine 50 mg 30 minutes prior to paclitaxel in order to avoid hypersensitivity reactions. Cisplatin 50 mg/m 2 was added on days 8 and 29 to the 24 h-infusion 5-FU backbone of the regimen. All patients received adequate antiemetic pre-medication prior to chemotherapy. A permanent venous access (PORT-A-Cath, Baxter Inc Germany) was implanted in all patients in order to facilitate the continuous 5-FU application. Portable single use 24 h infusion pumps (Infusor LV 5®, Baxter Inc, Germany) were used for the ambulatory application of 5-FU. All patients were treated on an outpatient basis.

Response and toxicity evaluation
Complete blood count, history and physical examination were recorded weekly. Renal and liver function tests were assessed biweekly in order to evaluate toxicity. Prior to each cycle, creatinine clearance, audiogram and electrocardiogram were rechecked. Toxicity was recorded every week based on WHO Toxicity Criteria. Dose modifications and treatment delays were performed as necessary according to the extent of the haematological and clinical toxicity. Planned dose modifications included the following: 25% dose reduction of paclitaxel and cisplatin in case of WHO grade 2 peripheral neurotoxicity; stop of paclitaxel and cisplatin in case of WHO grade 3/4 peripheral neurotoxicity; a 25% dose reduction of paclitaxel and/or 5-FU in case of WHO grade 3/4 granulocytopenia; 25% dose reduction of all 3 drugs in case of mucositis WHO grade ≥ 2. A 1-week treatment delay was planned in case of a WHO grade 3-4 toxicity, that had not resolved by the next week of treatment. Growth factors were not routinely used in the present study.
Baseline assessment of subjective symptoms (abdominal pain, vomiting, pain, loss of appetite, loss of weight) was performed with a modified EORTC-questionnaire by the patients themselves and their changes in intensity were recorded at the beginning, in the middle and at the end of each cycle. Patients were asked to grade their symptoms (absent, low, moderate, or severe) as well as their overall health status and quality of life on a scale reaching from 0 (very bad) to 100 (excellent). Questionaires were delivered by the treating physician and were to be completed prior to chemotherapy application.
Assessment of measurable disease was performed after each cycle according to WHO criteria (Miller et al, 1981). Complete remission (CR) was defined as the complete disappearance of all clinical, radiological and biochemical evidence of the disease and partial response as a greater than 50% reduction of all measurable tumour lesions lasting for at least 4 weeks. Stable disease was considered for all patients who achieved less than a partial response but no evidence of progressive disease (new lesions or increase in any area of measurable disease >25%). Treatment was continued if the patient showed a remission or stable disease. A maximum of 3 treatment cycles were planned per patient.

Study endpoints
Objectives of the present study were the determination of the overall (complete and partial) response rate, the median progression-free survival (PFS), the overall survival (OS) and the toxicity of the treatment. PFS, follow-up duration and OS were calculated from the start of treatment to the date of disease progression or the date of the last evaluation or death, respectively. Survival curves were estimated by the method of Kaplan-Meier (Kaplan and Meier, 1958). All statistics was performed using SAS software (SAS Institute, Cary, NC, version 6.11). Quality of life questionaires were descriptively analysed.

Forty-five patients with a median age of 60 years [35-75 years]
were enrolled in the study. Patient characteristics are given in Table 1. Tumour at the primary site, lymph nodes and liver metastases as well as peritoneal metastases were the most common tumour sites and all patients had bidimensionally measurable tumour lesions. Fourteen patients had undergone gastrectomy prior to chemotherapy and 2 patients had received a gastroenterostomy.

Response and survival
All patients were assessable for response and survival. Overall, 96 complete treatment cycles were administered to the 45 patients (median 3 cycles per patient; range 0-3 cycles). Five patients (11%) did not receive one complete treatment cycle for the following reasons: one patient refused further therapy due to WHO grade III°nausea and vomiting; one patient died of tumour bleeding following the second week of therapy; 3 patients were removed from therapy due to 5-FU related central neurotoxicity (1 patient), 5-FU-related cardiotoxicity with angina pectoris (1 patient) and a severe anaphylactic reaction to paclitaxel (1 patient). These patients were considered as treatment failures. Twenty-three patients [51%; 95% CI: 35.8-66.3.0%] achieved an objective response including 5 complete remissions (11%). Nine patients (20%) had stable disease as the best response to therapy and the remaining 8 patients (18%) developed progressive disease during treatment. Taking only the 40 patients into account, who received at least one complete cycle of therapy, the overall response rate was 58% [95% CI: 40.9-73.0%]. Responses occurred mainly at the primary tumour site and at liver, lymph node and lung metastases. The first patient with a complete remission presented with tumour at the primary site, liver metastases, an adrenal metastasis as well as a soft tissue metastasis. Following palliative gastrectomy and 3 cycles of chemotherapy complete remission was proven by CT scan. In two patients a complete remission, as confirmed by CT scan, was obtained after 3 cycles of chemotherapy for tumour at the primary site and liver metastases. One of these patients had had a palliative partial gastrectomy prior to chemotherapy. Two further patients with metastatic disease to the liver and to the lymph nodes and peritoneum had undergone a gastrectomy prior to chemotherapy. Both subsequently received 3 cycles of chemotherapy and a complete remission was confirmed by CT-scan. Two of these 5 patients relapsed, both at their initial disease sites. The durations of complete responses are currently 8+, 13, 18+, 36 and 36+ months.
Interestingly, 7 of 15 assessable patients with, among others, peritoneal metastases achieved a remission in the peritoneum, two patients a complete and five patients a partial remission. All remission evaluations were based on CT scan (not confirmed by laparoscopy). After a median follow-up of 14 months [1-36+], 17 patients have died of their disease and twenty-three patients are alive with 6 of them having progressive disease. Seven patients received a second-line chemotherapy, 4 of them mitomycin C (Hartmann et al, 1999). The median progression-free and overall survival are 9 (range 1-36+ months; 95% CI: 7-11) and 14 months (range 2-36+; 95% CI: 7-21), respectively.
Quality of life questionnaires were obtained from 32 patients and revealed an improvement in symptoms and overall well being in 14 patients (44%) and a subjectively stable condition in 12 patients (38%). Only 6 patients (19%) complained about a worsening of symptoms and physical well-being during therapy. Improvement of symptoms and overall well-being was clearly associated with tumour response. Twenty patients with an objective response and 6 patients with stable disease reported an improvement or stable status following therapy. Progressive disease was accompanied by a deterioration of symptoms in 4 patients. Only 1 patient who had obtained a complete remission complained about a worsening of symptoms and 1 patient with progressive disease reported an improvement of symptoms.

Toxicity
Toxicity according to WHO criteria was assessable in all 45 patients and is listed in Table 3 as the worst toxicity per patient during the total study period. Seven patients (15%) developed a WHO grade III/IV neutropenia. Based on the total number of cycles completed, neutropenia occurred in 17 of 96 cycles (18%). The duration of neutropenia was generally short and resolved   within 5 days. Only one patient had to be admitted to the hospital for WHO grade III fever and infection. This patient quickly responded to antibiotics without further complications, although the focus of the infection could not be detected. There was no WHO grade III°/IV°thrombocytopenia. The major WHO grade III/IV non-haematological toxicities were alopecia (71%), nausea/vomiting (11%) and mucositis (4%). No severe hand-foot syndrome was observed. In five patients treatment had to be stopped within the first cycle of therapy due to acute toxicities as stated above. Dose reductions (7 patients; 16%) and/or treatment delays (1 patient; 2%) were necessary in 8 patients (18%), in 7 patients due to neutropenia WHO grade III/IV and in 1 patient due to WHO grade III diarrhea. No dose reduction was performed due to neurotoxicity.

DISCUSSION
Second generation protocols for the treatment of advanced gastric cancer are mainly based on 5-FU, high-dose MTX, cisplatin and antracyclines. In phase II trials, response rates up to 60% have been reported for regimens such as FAMTX, ELF, EAP, Cisplatin/5-FU or ECF (Klein et al, 1986;Preusser et al, 1989;Wilke et al, 1990;Findley and Cunningham, 1993;Wils, 1996). However, in randomised phase III trials, this high level of activity has only been confirmed for the ECF-regimen (Webb et al, 1997;Waters et al, 1999), whereas for the FAMTX, ELF or cisplatin/5-FU-regimen response rates of 20-25% have been reported (Kelsen et al, 1992;Wilke et al, 1995). In addition, particularly the FAMTX and EAP regimens were associated with severe toxicity. Thus, to date, ECF appears to be the most active regimen, but a definitive standard regimen for the palliative treatment of metastatic gastric cancer has not yet been defined.
Several new agents have been tested in gastric cancer in the past years including topoisomerase I inhibitors such as CPT-11 or taxanes such as paclitaxel. In vitro and in vivo data appear to support the use of paclitaxel in gastric cancer (Chang et al, 1996;Ajani et al, 1998;Cascinu et al, 1998). An additive cytotoxic effect has been reported in vitro for the sequence of paclitaxel followed by 5-FU whereas the exposure to 5-FU followed by paclitaxel showed subadditive effects (Kano et al, 1996). Based on these rationale we had previously performed a phase II trial examining the efficacy and toxicity of the combination of paclitaxel, 5-FU and folinic acid in 22 patients with chemonaive gastric cancer (Bokemeyer et al, 1997a). A response rate of 32% and a median overall survival of 11 months were achieved. Toxicity was low with neutropenia WHO grade III/IV occurring in 14% of patients. In the study presented here as well as in our previous study, 5-FU was administered as a protracted, weekly 24 h infusion combined with folinic acid, since this mode of application appears to be less toxic and potentially more active as compared to the bolus administration. This regimen was initially investigated in patients with colon cancer and showed a good activity with low toxicity (Ardalan et al, 1991). In gastric cancer, one study had reported the use of protracted 24 h continuous infusion of high-dose 5-FU achieving a remission rate of 24% and a stable disease rate of 59% in patients previously treated with bolus 5-FU therapy . Thus, the weekly administration of 24 h continuous infusion of high-dose 5-FU (2000 mg/m 2 ) preceded by folinic acid (500 mg/m 2 as 2-h infusion) has formed the backbone for the development of our treatment regimen. Paclitaxel was added at a standard-dose of 175 mg/m 2 on days 1 and 22.
The present study investigates the efficacy and toxicity of the same regimen of paclitaxel and weekly 24-hour continuous infusion 5-FU plus folinic acid, but with the addition of cisplatin. Cisplatin has been shown to be active in gastric cancer as a single agent as well as in combination with 5-FU (Rougier et al, 1994;Schipper and Wagener, 1996;Wilke et al, 1996b). Furthermore, a synergistic effect has been described not only for the combination of 5-FU and cisplatin (Schabel et al, 1979;Scanlon et al, 1986) but also for the combination of paclitaxel and cisplatin in human gastric cancer cell lines . In order to avoid intensive toxicity in palliatively treated patients, cisplatin was given in a dose of 50 mg/m 2 on days 8 and 29. An objective response rate of 51% -or 58% if only those patients receiving at least one complete cycle of therapy are taken into accountincluding 11% (13%) complete remissions in chemonaive gastric cancer patients demonstrates the high activity of this regimen. A similar favourable response rate of 51% with a median survival time of 6 months was recently achieved by Kim et al (1999) administering paclitaxel at a dose of 175 mg/m 2 on day 1, 5-FU at a dose of 750 mg/m 2 on days 1-5 as continuous infusion and cisplatin at a dose of 20 mg/m 2 as a 30 minute infusion on days 1-5. In patients with adenocarcinoma of the oesophagus, Ilson et al (1998) reported a response rate of 46% with a similar schedule. The median overall and progression-free survival of 14 and 9 months duration, respectively, in the current study are encouraging. The addition of cisplatin to our previous regimen consisting of paclitaxel and 5-FU/folinic acid appears to improve the overall response rate, in particular the number of complete responses and the median survival time (Bokemeyer et al, 1997b).
The toxicity of the present combination of paclitaxel, 24 h continuous infusion of high-dose 5-FU/folinic acid and cisplatin was acceptable and the treatment could be safely administered on an outpatient basis. Only one patient had to be hospitalized due to fever and infection. Dose reductions or treatment delays were performed in 18% of patients, mainly due to neutropenia. More than 60% of the patients who participated in the quality of life evaluation reported an improvement or stabilization of their symptoms and general condition. This clinical benefit not only reflects the favourable toxicity profile but also demonstrates the palliative effect of this regimen.  However, results of phase II trials have to be carefully interpreted since patient selection may cause a bias and subsequently misleading results. Thus, in order to confirm the high activity of the current regimen and in order to define the role of paclitaxel, randomized phase III trials are needed. The EORTC study 40953 has compared ELF to FAMTX and to 5-FU/folinic acid as well as to 5-FU/folinic acid plus cisplatin. Results of this study may yield a control arm for future randomized trials. Due to its high response rate and its moderate toxicity allowing outpatient treatment, the combination of high-dose 5-FU/folinic acid, cisplatin plus paclitaxel may be an attractive option to be tested in neoadjuvant or adjuvant trials in patients with locally advanced disease.