A phase III trial of neoadjuvant intraperitoneal and systemic chemotherapy for gastric cancer with peritoneal metastasis

Gastric cancer (GC) is one of the most prevalent malignancies worldwide, with nearly 1 million new cases each year [1]. In contrast to countries with national screening programs, such as Japan and Korea, GC in China is often diagnosed at an advanced stage, as early-stage tumors are usually asymptomatic [2]. Patients with advanced GC often develop recurrence after curative resection, and in these cases peritoneal dissemination is the most common pattern of recurrence [3]. It has been reported that about 10–20% of patients with T3 or T4 GC already have microscopic peritoneal metastases that are difficult to identify when radical surgery is performed [4]. Once peritoneal metastases have progressed, they can eventually result in intractable ascites, intestinal obstruction and cachexia, which are the main causes of death in GC. Peritoneal recurrence is the most important reason for treatment failure in GC patients, and it develops in up to 70% of patients and has an average survival of merely 6 months after diagnosis [5]. This study is being conducted to investigate the feasibility and efficacy of more aggressive treatment for GC patients with peritoneal metastasis.

DRAGON-01 study

DRAGON-01 is a multicenter randomized controlled trial that will be carried out in multiple hospitals throughout China. Eligible patients with GC with peritoneal metastasis will be randomized at a ratio of 2:1 into two groups to receive an intraperitoneal and intravenous paclitaxel plus S-1 regimen (PS; neoadjuvant intraperitoneal–systemic [NIPS] chemotherapy) or intravenous paclitaxel plus S-1 regimen. The study was approved by the ethics committee of Ruijin Hospital, Shanghai Jiao Tong University School of Medicine, Shanghai, China, and was registered 6 November 2016, in the WHO International Clinical Trials Registry Platform, version 1.0, with the registration identifier ChiCTR-IIR-16009802. All patients entering the study will be required to sign informed consent. This work underwent peer review and is supported by the Multicenter Clinical Research Project of Shanghai Jiao Tong University School of Medicine (project number: DLY201602). This work is also supported by the National Natural Science Foundation of China (project number: 81772518). The funding sources had no role in the design of the study and will not have any role during the execution, analysis, interpretation of data or decision to submit results. The purpose of the funding is to develop and sustain the implementation of the multicenter randomized controlled trial, including the third-party specialized quality inspection. Monitoring will be carried out throughout the trial. The participating centers are Ruijin Hospital, Shanghai Changhai Hospital, Zhejiang Provincial People's Hospital, The First Affiliated Hospital of Zhejiang University Medical College and The Second Affiliated Hospital of Zhejiang University Medical College and Shandong Cancer Hospital.

Background & rationale

Although recent advances in systemic chemotherapy have improved the clinical outcomes of GC patients with peritoneal metastasis, these patients still have a poor prognosis [6]. Over the past 2 decades, a multimodal treatment utilizing cytoreductive surgery plus neoadjuvant chemotherapy was developed, which takes advantage of surgery to reduce visible tumor burden and neoadjuvant chemotherapy to eradicate peritoneal micrometastases and free cancer cells [7]. However, peritoneal penetration of conventional chemotherapy is poor, and only a small fraction of the systemically administered drug is delivered to the peritoneum. To overcome this limitation, a new bidirectional chemotherapy (NIPS) was developed to induce a reduction in peritoneal cancer index of peritoneal carcinomatosis due to GC. NIPS chemotherapy can attack peritoneal carcinomatosis from the peritoneal cavity and subperitoneal blood vessels [8].

To date, there are preliminary reports regarding the effectiveness of NIPS chemotherapy in GC with peritoneal metastasis. Yonemura et al. reported on a multidisciplinary treatment combined with NIPS chemotherapy (60 mg/m² of oral S-1 for 21 days followed by 1 week of rest, with 30 mg/m² of Taxotere and 30 mg/m² of cisplatin introduced through the port on days 1, 8 and 15 and cytoreductive surgery performed after two cycles of NIPS chemotherapy) that was used in 96 GC patients with peritoneal carcinomatosis [7]. In this study, 68 patients showed positive cytology before NIPS chemotherapy, and the positive cytology results became negative in 47 (69%) patients after NIPS chemotherapy. Complete cytoreduction was achieved in 58 (70.7%) patients, and significant improvement in survival was indicated by multivariate analysis. Results of the phase III PHOENIX-GC trial also demonstrated the promising clinical effects of NIPS chemotherapy [9]. In this randomized phase III trial, 164 patients with GC with peritoneal metastasis were randomly assigned at a 2:1 ratio to receive intraperitoneal and intravenous paclitaxel plus S-1 (IP; intraperitoneal paclitaxel 20 mg/m² and intravenous paclitaxel 50 mg/m² on days 1 and 8 plus S-1 at 80 mg/m² per day on days 1–14 for a 3-week cycle) or S-1 plus cisplatin (SP; S-1 at 80 mg/m² per day on days 1–21 plus cisplatin 60 mg/m² on day 8 for a 5-week cycle). The median survival time (MST) was 17.7 months in the IP arm and 15.2 months in the SP arm (hazard ratio: 0.72; p = 0.080), and the 3-year overall survival (OS) rate was 21.9% in the IP arm and 6.0% in the SP arm. Although this trial failed to show statistical superiority of NIPS chemotherapy, the exploratory analyses suggested possible clinical benefits of the NIPS chemotherapy protocol.

In addition, several recent studies have indicated that intraperitoneal administration of anticancer drugs is a reasonable method for treating peritoneal metastasis [10–13]. According to these studies, up to 41.9–78.0% of patients were able to undergo conversion surgery, and the rates of R0 resection were 55.6–68.8%, with postoperative MST of 12.8–43.2 months and 3-year OS of 12.0–55.3%. By contrast, the MST of patients unable to undergo conversion surgery was 8.0–10.3 months. Therefore, NIPS chemotherapy plus conversion surgery may be an appropriate curative treatment for GC patients with peritoneal metastasis.

Based on these previous results, the authors planned a prospective randomized controlled phase III trial (DRAGON-01) to investigate the feasibility and efficacy of more aggressive treatment for GC patients with peritoneal metastasis. The results of Japanese clinical studies have suggested that NIPS chemotherapy might be potentially effective in improving the prognosis of these patients, but there is still lack of evidence regarding this new technique, especially in China [9]. In the PHOENIX-GC trial, paclitaxel was used intravenously in the IP group, whereas cisplatin was used in the SP group. Therefore, it is difficult to distinguish whether the difference in efficacy between the two groups was due to a difference in the drug or a difference in the route of administration. Moreover, the gastrointestinal toxicity of cisplatin in the Cisplatin + S-1 (CS) chemotherapy regimen was comparatively notable in PHOENIX-GC trial. Patients with advanced GC are usually characterized by poor fitness and poor tolerance of chemotherapy.

To overcome these disadvantages, the authors have decided to use a PS chemotherapy regimen in the control group so that only the route of administration differs between the NIPS and PS chemotherapy arms. According to previous studies, the PS chemotherapy regimen is equivalent to the Cisplatin + S-1 (CS) chemotherapy regimen in terms of efficacy for unresectable or recurrent GC and has a lower incidence of gastrointestinal adverse effects such as anorexia, nausea and vomiting [14]. The Ruijin Hospital Ethics Committee approved this study in November 2016, and the study was registered in the Chinese Clinical Trial Registry as ChiCTR-IIR-16009802. This ongoing study has not completed participant recruitment at the time of submission, and over 220 patients have been recruited in six participating centers.

Study design

DRAGON-01 is a prospective multicenter randomized controlled phase III trial (Figure 1). The authors plan to enroll a total of 238 patients over 6 years.

Preoperative staging

Patients are evaluated with the following examinations to ascertain whether peritoneal metastasis is suspected and to determine that there is no evidence of other distant metastasis: complete blood count and comprehensive chemistry profile; tumor markers (AFP, carcinoembryonic antigen, CA125, CA19–9, CA724); contrast-enhanced thoracic, abdominal and pelvic computed tomography (CT); and PET/CT if no evidence of distant metastasis and if clinically indicated.

Registration

For patients who give consent and fulfill the inclusion criteria, diagnostic laparoscopy is routinely performed to confirm the presence of peritoneal metastasis and to evaluate the peritoneal cancer index introduced by Jacquet and Sugarbaker [15]. After the laparoscopic exploration confirming peritoneal carcinomatosis, randomization is carried out by an online central randomization service that generates allocation in a 2:1 design protocol. According to study design, patients are separated into two arms: NIPS chemotherapy and PS chemotherapy.

NIPS chemotherapy arm

In the NIPS chemotherapy arm, a subcutaneous intraperitoneal chemotherapy pump is placed in the lower abdomen or an intraperitoneal tube is placed by other means, preferably in the right lower abdomen, with the catheter placed in the pelvic floor. Patients receive intraperitoneal and intravenous paclitaxel plus oral S-1.

PS chemotherapy arm

The PS chemotherapy arm comprises the control group, and patients receive intravenous paclitaxel plus oral S-1.

Neoadjuvant chemotherapy

The NIPS chemotherapy regimen comprises intraperitoneal paclitaxel 20 mg/m² and intravenous paclitaxel 50 mg/m² on days 1 and 8 plus oral S-1 at 80 mg/m² per day (40–60 mg twice per day, depending on body surface area, as follows: <1.25 m², 40 mg; 1.25–1.5 m², 50 mg; >1.5 m², 60 mg) on days 1–14 of every 3-week cycle. A total of 50 mg/m² paclitaxel is diluted in 500 ml of normal saline and administered intravenously, and 20 mg/m² paclitaxel is diluted in 1000 ml of normal saline and administered intraperitoneally via an intraperitoneal port or catheter. The PS chemotherapy regimen comprises intravenous paclitaxel 70 mg/m² on days 1 and 8 plus oral S-1 at 80 mg/m² per day on days 1–21 of every 3-week cycle. A total of 70 mg/m² paclitaxel is diluted in 500 ml of normal saline and administered intravenously.

The NIPS and PS chemotherapy regimens are usually started the day after laparoscopic exploration. Treatment with either regimen is terminated if disease progression or intolerable adverse effects are assessed.

Assessment of tumor response

Patients in both arms are evaluated every three courses to verify the safety and curative effect of treatments, including general condition, tumor markers, amount of ascites and changes in primary lesion and peritoneal metastasis. A second-look laparoscopic exploration will be performed to determine whether conversion surgery is indicated. If the patient does not meet the indication for conversion surgery, the course of treatment will be continued for another two to three courses, after which the patient will be reevaluated.

Conversion surgery

The indication for conversion surgery includes resectable primary tumor, disappearance of peritoneal metastasis, no free cancer cell detected, no distant metastases other than peritoneal metastases, aforementioned requirements confirmed by CT scan and laparoscopic exploration and clinical benefits, including increased appetite and improved abdominal symptoms and nutritional status. Oral S-1 from the last course needs to be stopped for 2–3 weeks before conversion surgery. A standard R0 open gastrectomy with D2 lymphadenectomy is recommended. Laparoscopic adnexectomy is preferred immediately after a second laparoscopy if adnexal metastases are confirmed. The type of gastrectomy is determined according to recommendations by authoritative guidelines, such as Japanese gastric cancer treatment guidelines [17] or National Comprehensive Cancer Network guidelines [18]. The type of gastrectomy is determined according to recommendations by authoritative guidelines [16–18]. Combined organ resection, such as splenectomy and colectomy, is allowed in order to facilitate dissection of metastatic lymph nodes in the relevant areas. D0 or D1 lymphadenectomy can be chosen for palliative surgery if necessary for tumor debulking.

Adjuvant chemotherapy

The regimen for adjuvant chemotherapy is the same as that used for neoadjuvant chemotherapy. The status and recovery of patients will be assessed at a time point set at around 3 weeks postoperatively. The postoperative course of treatment will be scheduled depending on the status of the patient. In case of disease progression or intolerable adverse effects during treatment, the patient will be referred to a multidisciplinary team for further assessment of the course of treatment. Depending on the circumstances, second-line chemotherapy, best supportive care or follow-up may be indicated.

Toxicity criteria

Adverse events will be assessed according to Common Terminology Criteria for Adverse Events, version 4.0.

NIPS chemotherapy arm

For severe digestive adverse reactions such as anorexia, nausea and diarrhea, reduce S-1 by 20 mg/m². For grade 3/4 leukopenia or neutropenia, reduce intravenous paclitaxel by 5 mg/m². If abdominal pain develops after intraperitoneal injection (excluding acute abdomen), reduce intraperitoneal paclitaxel by 5 mg/m². Further increments will not be considered in subsequent cycles after dose reduction. Patients need to recover to grade 0/1 before proceeding to the next cycle, which may be delayed by up to 3 weeks.

PS chemotherapy arm

S-1 or paclitaxel may be reduced by 10 mg/m² according to the adverse effects of the previous course. S-1 is not normally considered for dose reduction within the same course of treatment except for severe adverse effects.

Follow-up

Follow-up of all patients will be carried out before each course of treatment. Physical examination, tumor marker examination and abdominal CT are performed every three courses. Disease progression is defined as any direct signs (e.g., peritoneal thickening, nodular changes, mesenteric infiltration or pelvic masses) or indirect signs (e.g., ascites, retroperitoneal lesions, renal effusion or ureteral obstruction due to compression from recurrent masses) of local recurrence found on CT or signs of local recurrence discovered during endoscopic examination.

Eligibility criteria

The eligibility criteria of this study are as follows:

• Histologically confirmed GC without prior history of gastric malignancy;

• Peritoneal metastasis from GC requiring definitive diagnosis by laparoscopy without gastric outflow tract obstruction or intestinal obstruction;

• No prior treatment of chemotherapy, radiation therapy, targeted therapy or immunotherapy;

• Age between 18 and 75 years at registration;

• Eastern Cooperative Oncology Group score ≤2;

• Expected life expectancy ≥3 months;

• Adequate organ function to tolerate surgery and chemotherapy, including:

  • Normal bone marrow function (neutrophils ≥1.5 × 10⁹/l, hemoglobin ≥8.0 g/dl, platelets ≥100 × 10⁹/l);

  • Normal liver function (total bilirubin ≤1.5 mg/dl, AST ≤100 IU/l, ALT ≤100 IU/l);

  • Normal kidney function (creatinine clearance >50 ml/min).

Exclusion criteria

The exclusion criteria of this study are as follows:

• Confirmed evidence of distant metastasis other than peritoneal metastasis (e.g., liver metastasis, lung metastasis or para-aortic lymph node metastasis);

• Currently pregnant, within 28 days of parturition, or lactating;

• Synchronous or metachronous (within 5 years) malignancies;

• Severe mental disease, uncontrolled epilepsy or CNS disease;

• Clinically severe (i.e., active) heart disease, such as symptomatic coronary artery disease, New York Heart Association class II or more severe congestive heart failure or arrhythmia requiring drug intervention or a history of myocardial infarction in the last 12 months;

• Upper gastrointestinal obstruction or abnormal physiological function or malabsorption syndrome that may affect S-1 absorbers;

• Known peripheral neuropathy (however, patients with only disappearance of deep tendon reflex need not be excluded);

• Patients on steroid or immunosuppressive treatment after organ transplant;

• Patients with severe uncontrolled recurrent infections or other severe uncontrolled concomitant disease;

• Moderate or severe renal damage (creatinine clearance ≤50 ml/min) or serum creatinine greater than upper limit of normal;

• Known DPYD deficiency;

• Anaphylaxis to paclitaxel or any research drug ingredient.

Endpoints

The primary endpoint is OS. OS is calculated from the date of randomization to the date of death or date of last follow-up. The secondary endpoints are pathological response rate, gastrectomy radicality rate, progression-free survival and adverse events. Pathological response rate is defined as the proportion of patients achieving tumor regression grade 0–2 according to the histological criteria of the Chinese Society of Clinical Oncology. Progression-free survival is calculated from the date of randomization to the date of detected disease recurrence. The following events are defined as recurrence: primary cancer recurrence, newly diagnosed GC and death. Adverse events during pre- and postoperative chemotherapy are evaluated according to the Common Terminology Criteria for Adverse Events, version 4.0. The Clavien–Dindo classification is used to describe postoperative adverse events [19]. Port-related complications are also collected as adverse events in this study. Detailed background data, including comorbidities, past medical history, pathology results, surgery information and postoperative examinations, are collected during this trial. The authors will report the collected background data with the primary analysis results.

Statistics

The main endpoint of this study is OS. Based on the results of previous studies [20–23], the authors expect MSTs of 22 and 11 months in the NIPS and PS chemotherapy arms, respectively. The random distribution ratio of the experimental and control groups is 2:1, with a statistical test efficiency (1-β) of 80% and a one-sided test level of significance (α) of 0.1. The expected study subjects will be enrolled for 6 years and followed up for 3 years. Using PASS 11.0 software (NCSS, UT, USA) for sample size estimation, the log-rank test revealed a sample size of 216, which, when factoring in a dropout rate of 10%, was estimated to be around 238, with 158 patients in the NIPS chemotherapy arm and 80 patients in the PS chemotherapy arm.

With regard to primary and secondary endpoints, survival rates are analyzed using survival curves calculated by the Kaplan–Meier method, and differences between survival curves are examined with the log-rank test. Prognostic factors are analyzed using Cox regression. Clinicopathological variables are analyzed using chi-square test for discrete variables and t-test for continuous variables. Subgroup analyses will be performed to fairly evaluate the results according to age, sex, extent of peritoneal metastasis, amount of ascites, etc. A p < 0.05 is considered significant.

Conclusion

We have described the design of and rationale for a prospective multicenter randomized controlled phase III trial (DRAGON-01) being conducted to evaluate the feasibility and efficacy of NIPS chemotherapy treatment for GC patients with peritoneal metastasis. The high prevalence of peritoneal carcinomatosis in GC patients has prompted the search for more aggressive treatment options to manage the dissemination of cancer cells. This is the first prospective multicenter randomized study to investigate the role of NIPS chemotherapy in the management of GC patients with peritoneal metastasis in China based on an optimized regimen according to the Japanese PHOENIX-GC trial. If the survival rate in the NIPS chemotherapy arm is superior to that observed in the PS chemotherapy arm, this trial will provide favorable evidence for clinical guidelines when considering the preferred treatment for GC patients with peritoneal metastasis.