Effect of First-line S-1 Plus Oxaliplatin With or Without Ramucirumab Followed by Paclitaxel Plus Ramucirumab on Advanced Gastric Cancer in East Asia: The Phase 2 RAINSTORM Randomized Clinical Trial.

Importance: Ramucirumab, a human IgG 1 antibody against vascular endothelial growth factor receptor 2, has been shown to improve progression-free survival and overall survival in patients with advanced gastric cancer in the second-line setting. Objective: To compare progression-free survival for S-1 and oxaliplatin plus ramucirumab with that for S-1 and oxaliplatin plus placebo in patients with advanced gastric cancer. Design, Setting, and Participants: This phase 2, double-blind randomized clinical trial (RAINSTORM [First-line S-1 Plus Oxaliplatin With or Without Ramucirumab Followed by Paclitaxel Plus Ramucirumab in Patients With Advanced Gastric Cancer]) was conducted from October 12, 2015, to April 11, 2018, at 36 sites in Japan, South Korea, and Taiwan. Participants were chemotherapy-naive patients (n = 189) with metastatic gastric or gastroesophageal adenocarcinoma. Analyses of the full analysis set and safety population were conducted between November 27, 2017, and June 4, 2018. Interventions: Patients randomized to the ramucirumab plus S-1 and oxaliplatin arm received S-1, 80 to 120 mg/d twice daily, on days 1 to 14 and oxaliplatin, 100 mg/m2, on day 1 with ramucirumab, 8 mg/kg, on days 1 and 8 in part A (21-day cycle). Patients randomized to the placebo plus S-1 and oxaliplatin arm received the same S-1 and oxaliplatin dosage as well as placebo on days 1 and 8 in part A. Eligible patients received second-line paclitaxel, 80 mg/m2, on days 1, 8, and 15 and ramucirumab, 8 mg/kg, on days 1 and 15 in part B (28-day cycle). Main Outcomes and Measures: The primary end point was progression-free survival, analyzed using the stratified log-rank test; the hazard ratio (HR) was estimated using the stratified Cox proportional hazards regression model. Secondary end points included overall survival and adverse events.
Results: In total, 189 patients were randomized and received treatment: 96 to the ramucirumab plus S-1 and oxaliplatin arm and 93 to the placebo plus S-1 and oxaliplatin arm. Among the 189 patients, 121 (64.0%) were male, and the median (range) age was 62.0 (26-84) years. Median progression-free survival was not prolonged in the ramucirumab plus S-1 and oxaliplatin arm compared with the placebo plus S-1 and oxaliplatin arm (6.34 [80% CI, 5.65-6.93] vs 6.74 [80% CI, 5.75-7.13] months; HR, 1.07; 80% CI, 0.86-1.33; P = .70). Median overall survival was 14.65 (80% CI, 12.39-15.67) months in the ramucirumab plus S-1 and oxaliplatin arm and 14.26 (80% CI, 13.83-17.31) months in the placebo plus S-1 and oxaliplatin arm (HR, 1.11; 80% CI, 0.89-1.40; P = .55). The most commonly reported grade 3 or higher treatment-emergent adverse events in the ramucirumab plus S-1 and oxaliplatin arm in part A were decreased neutrophil count (14 patients [14.6%]), hypertension (10 patients [10.4%]), and anemia (10 patients [10.4%]). Conclusions and Relevance: In this randomized clinical trial, the addition of ramucirumab to first-line S-1 and oxaliplatin treatment did not prolong progression-free survival or overall survival compared with S-1 and oxaliplatin alone among East Asian patients with advanced gastric cancer; no new safety signals for ramucirumab were identified. Trial Registration: ClinicalTrials.gov identifier: NCT02539225.

RCT Entities:   Population Interventions Outcomes

Introduction

Globally, gastric cancer is the fifth most frequently diagnosed cancer and the third leading cause of cancer death.[1] More than half of the cases in the world occur in East Asia.[2] Chemotherapy is the main treatment option for patients with advanced gastric or gastroesophageal junction adenocarcinoma.[3,4] The current standard of care in the first-line setting is combination chemotherapy with a fluoropyrimidine (eg, fluorouracil, S-1 [tegafur-gimeracil-oteracil potassium], capecitabine) plus a platinum agent (eg, cisplatin, oxaliplatin).[3,4] Among these chemotherapy regimens, S-1 plus oxaliplatin is frequently used in East Asia, according to the phase 3 G-SOX study.[5] However, as first-line chemotherapy is associated with median overall survival (OS) times in the range of 6 to 13 months in clinical trials,[4] other treatment options with improved clinical outcomes are required for patients with advanced gastric or gastroesophageal junction adenocarcinoma.

Ramucirumab is a human IgG 1 monoclonal antibody against vascular endothelial growth factor (VEGF) receptor 2 that blocks the binding of VEGF ligands, thereby preventing receptor activation and inhibiting angiogenesis.[6] Ramucirumab has shown promising activity in the treatment of advanced gastric or gastroesophageal junction adenocarcinoma in the second-line setting as monotherapy or in combination with paclitaxel.[7,8] In particular, paclitaxel plus ramucirumab has become the standard second-line chemotherapy for advanced gastric or gastroesophageal junction adenocarcinoma in East Asia, where paclitaxel monotherapy was a community standard.[4] However, the role of an antiangiogenic agent when added to chemotherapy in the first-line setting remains unclear,[9,10,11] including the addition of ramucirumab to the global standard of care of fluorouracil-cisplatin or capecitabine-cisplatin or the East Asian standard of care of S-1 plus oxaliplatin. Furthermore, investigation of the continuous use of ramucirumab from the first-line setting to the second-line setting would be of interest, given the finding that the continued use beyond disease progression of bevacizumab (another angiogenesis inhibitor) while switching chemotherapy is beneficial in patients with metastatic colorectal cancer.[12]

Two clinical trials have examined these applications of ramucirumab in patients with metastatic gastric or gastroesophageal junction adenocarcinoma: (1) the global, randomized, phase 3 RAINFALL (Ramucirumab With Cisplatin and Fluoropyrimidine as First-line Therapy in Patients With Metastatic Gastric or Junctional Adenocarcinoma) study,[13] in which the addition of ramucirumab to fluorouracil-cisplatin or capecitabine-cisplatin was evaluated in the first-line setting; and (2) the East Asian, randomized, phase 2 RAINSTORM (First-line S-1 Plus Oxaliplatin With or Without Ramucirumab Followed by Paclitaxel Plus Ramucirumab in Patients With Advanced Gastric Cancer) study described here, in which the addition of ramucirumab to S-1 plus oxaliplatin in the first-line setting (part A) and paclitaxel plus ramucirumab treatment in the second-line setting (part B) were evaluated. Given that potential ethnicity-associated differences in efficacy have been observed in the AVAGAST (Avastin in Gastric Cancer)[9] and RAINBOW (Ramucirumab Plus Paclitaxel vs Placebo Plus Paclitaxel in Patients With Previously Treated Advanced Gastric or Gastroesophageal Junction Adenocarcinoma)[8] studies and S-1 plus oxaliplatin is becoming a mainstay of first-line treatment in Japan, we set up the RAINSTORM study in parallel with the RAINFALL study.

The primary objective of the RAINSTORM study was to evaluate progression-free survival (PFS). The dose of ramucirumab in part A of this study (8 mg/kg on days 1 and 8 of a 21-day cycle) differed from the approved dose (8 mg/kg every 2 weeks).[14] This 3-week schedule was chosen to achieve a higher ramucirumab exposure, on the basis of modeling kinetics that suggested an exposure-response relationship[15] and the phase 1b JVCX (Phase 1b Study of Ramucirumab in Combination With Fluoropyrimidines and Platinum-Based Agents in Japanese Patients With Metastatic Gastric/Gastroesophageal Junction Adenocarcinoma) study, conducted in chemotherapy-naive Japanese patients with metastatic gastric or gastroesophageal junction adenocarcinoma, which showed that higher trough levels of ramucirumab were achieved with this schedule without notable increases in toxic effects.[16]

Methods

Study Design

This trial was a randomized, double-blind, phase 2 study of S-1 plus oxaliplatin with or without ramucirumab among East Asian patients with metastatic gastric or gastroesophageal junction adenocarcinoma. It was conducted between October 12, 2015, and April 11, 2018, at a total of 36 sites, with 25 sites in Japan, 6 in South Korea, and 5 in Taiwan. The analysis was performed from November 27, 2017, to June 4, 2018. The trial protocol (Supplement 1) was approved by each site’s ethics review board. All participants provided written informed consent before undergoing any protocol-specific evaluations. The study was conducted in accordance with the Declaration of Helsinki,[17] Good Clinical Practice guidelines, and applicable local regulations and followed the Consolidated Standards of Reporting Trials (CONSORT) reporting guideline.

Study Population

Patients with histologically or cytologically confirmed metastatic gastric or gastroesophageal junction adenocarcinoma who had not received first-line systemic therapy for metastatic gastric or gastroesophageal junction adenocarcinoma (previous neoadjuvant or adjuvant therapy was permitted) were eligible for this study (Figure 1). Patients with esophageal cancer were not eligible. The main inclusion criteria were age 20 years or older; measurable disease or nonmeasurable but evaluable disease according to the Response Evaluation Criteria in Solid Tumors (RECIST), version 1.1[18]; Eastern Cooperative Oncology Group Performance Status score of 0 or 1; and adequate hepatic, kidney, hematologic, and coagulation function.

Figure 1.

CONSORT Flow Diagram

At the time of data cutoff, 15 patients were still receiving treatment in part A, 2 patients were in the follow-up period after part A, and 17 patients were still receiving treatment in part B. The 2 patients in the ramucirumab plus S-1 and oxaliplatin (RAM+SOX) arm and the 1 patient in the placebo plus S-1 and oxaliplatin (PBO+SOX) arm who discontinued part A for the reason of death are not included in the numbers of patients after part A. In part B, all patients were to receive second-line paclitaxel plus ramucirumab (PTX+RAM) treatment.

The main exclusion criteria were ERBB2-positive status; significant bleeding disorders, vasculitis, or a significant bleeding episode from the gastrointestinal tract within 12 weeks before randomization; any arterial thromboembolic event within 24 weeks before randomization; and history of gastrointestinal disorders, including perforation and/or fistulae within 24 weeks before randomization, inflammatory bowel disease requiring medical intervention 48 weeks or less before randomization, acute or subacute bowel obstruction, or a history of chronic diarrhea.

Randomization and Masking

Patients were randomized 1 to 1 through an interactive web response system (IWRS) to receive either ramucirumab plus S-1 and oxaliplatin or placebo plus S-1 and oxaliplatin. Randomization was stratified by Eastern Cooperative Oncology Group Performance Status score (0 or 1), region (Japan or other [South Korea or Taiwan]), and disease measurability (measurable or nonmeasurable), with a block size of 4. The treatment arm information for the trial was generated on the basis of a seed number, IWRS chunk size, number of strata, and block size and was then passed on to the IWRS. The IWRS team generated the random allocation sequence, and the study sites were responsible for enrollment and randomization. Only the IWRS team and product delivery team had access to the unblinded information. The investigators, patients, and sponsor personnel remained blinded during part A until database lock for the primary end point analysis.

Treatment Protocol

Part A of the study compared first-line S-1 plus oxaliplatin with and without ramucirumab (eFigure 1 in Supplement 2). In part B, all eligible patients received second-line paclitaxel plus ramucirumab to avoid crossover administration of ramucirumab (ie, first-line ramucirumab to second-line placebo or first-line placebo to second-line ramucirumab).

Part A

Patients received intravenous ramucirumab, 8 mg/kg, or placebo on days 1 and 8 of a 21-day cycle, in combination with oral S-1, 80 to 120 mg, twice daily on days 1 to 14 and intravenous oxaliplatin 100 mg/m2 on day 1. Ramucirumab or placebo and S-1 plus oxaliplatin were continued until disease progression, unacceptable toxic effects developed, or any other discontinuation criteria were met.

Part B

Enrolled patients in both arms of part A who progressed and were eligible for second-line therapy were considered for part B. After the completion of assessments during the pretreatment period of part B, eligible patients received intravenous ramucirumab, 8 mg/kg, on days 1 and 15 plus paclitaxel, 80 mg/m2, on days 1, 8, and 15 every 28 days. Patients who did not meet the initiation criteria for part B (eFigure 1 in Supplement 2) within 12 weeks from the decision to discontinue study treatment in part A did not continue participation in the study.

Efficacy

The primary end point was PFS (first disease progression) in part A, measured from the date of randomization to the date of radiographic documentation of progression (as assessed by an investigator using the RECIST, version 1.1) or death due to any cause, whichever was earlier. Progression-free survival was censored (1) at the date of the most recent adequate radiologic assessment or the date of randomization (whichever was later) for patients without disease progression or who had not died; (2) at the date of randomization for patients who had missing baseline radiologic tumor assessment and for patients who had missing adequate postbaseline radiologic tumor assessment; (3) at the date of adequate radiologic assessment on or before new anticancer therapy or the date of randomization (whichever was later) for patients receiving new anticancer treatment, including the second-line therapy (paclitaxel plus ramucirumab); and (4) at the date of the most recent adequate radiologic assessment before the missing assessment or the date of randomization (whichever was later) for patients who progressed or for whom death was documented immediately after 2 or more consecutive missing scan intervals after the most recent adequate radiologic tumor assessment or randomization (whichever was later).

Secondary end points included second disease progression (PFS2), OS, overall response rate, and disease control rate. The PFS2 was measured from the date of randomization to the date of radiographic documentation of progression after the start of second-line therapy (paclitaxel plus ramucirumab) or death, using the tumor assessment before starting second-line therapy as the baseline assessment. If second-line therapy was not started, OS was substituted for PFS2. Overall survival was measured from the date of randomization to the date of death from any cause. Overall response rate was the proportion of randomized patients with measurable disease who achieved a best overall response of complete response or partial response in part A according to RECIST, version 1.1. Disease control rate was the proportion of randomized patients with measurable disease who achieved a best overall response of complete response, partial response, or stable disease in part A according to RECIST, version 1.1. Best overall response was the best response recorded from randomization until disease progression.

Pharmacokinetics and Safety

In part A, samples for determining ramucirumab serum trough concentration levels were scheduled before the ramucirumab dose on day 8 of cycle 1 and on day 1 of cycles 2, 3, 5, and 9. Adverse events (AEs) in part A and part B were graded according to the National Cancer Institute Common Terminology Criteria for Adverse Events, version 3.0.

Statistical Analysis

The target sample size was 190 patients and 136 PFS events for a power of 80% with a 2-sided significance level of .20, assuming a PFS hazard ratio (HR) of 0.67. The full analysis set for efficacy (based on randomized arm; baseline and efficacy analyses) and the safety population (based on actual arm; safety and exposure analyses) were defined as all randomized patients who received any study medication; in practice, the 2 populations were the same. The PFS was analyzed with the stratified log-rank test, and the PFS HR was estimated using the stratified Cox proportional hazards regression model; both were stratified by the randomization strata. P values were based on stratified log-rank tests. The same statistical methods were used for the analysis of PFS2 and OS. The forest plots (eFigure 2 in Supplement 2) were based on an unstratified Cox proportional hazards regression model for each subgroup. Odds ratios (ORs) for the overall response rate and disease control rate and their P values were based on logistic regression and the Cochran-Mantel-Haenszel test stratified by the randomization strata.

Tests for treatment effect were conducted at a 2-sided significance level of .20. Treatment-emergent AEs (TEAEs) were any untoward or worsened medical occurrence any time after baseline that did not necessarily have a causal relationship with the treatment. The AEs of special interest included arterial and venous thromboembolic events, bleeding or hemorrhage, fistula, gastrointestinal hemorrhage events, gastrointestinal perforation, hypertension, infusion-related reaction, proteinuria, and renal failure.

All results were from the primary database lock (November 27, 2017), except for the PFS2 and OS results, which were from the second database lock (June 4, 2018). Analyses of the full analysis set and safety population were performed using SAS, version 9.4 (SAS Institute).

Results

Patient Disposition

In total, 191 patients were randomized, of whom 2 did not receive treatment, resulting in 189 patients included in the full analysis set and the safety population in part A. Ninety-six patients were randomized to receive ramucirumab plus S-1 and oxaliplatin, and 93 patients were randomized to receive placebo plus S-1 and oxaliplatin (Figure 1). Among the 189 patients, 121 (64.0%) were male, and the median (range) age was 62.0 (26-84) years. Baseline characteristics were generally similar between the 2 arms in part A (Table 1).

Table 1.

Baseline Patient Characteristics in Part A

Variable	No. (%)
	RAM+SOX Arm (n = 96)	PBO+SOX Arm (n = 93)
Male	57 (59.4)	64 (68.8)
Age, median (range), y	61.0 (30-82)	63.0 (26-84)
<65	57 (59.4)	53 (57.0)
≥65	39 (40.6)	40 (43.0)
Region
Japan	65 (67.7)	64 (68.8)
Korea	20 (20.8)	20 (21.5)
Taiwan	11 (11.5)	9 (9.7)
ECOG PS
0	64 (66.7)	62 (66.7)
1	32 (33.3)	31 (33.3)
Primary tumor location
Gastric	91 (94.8)	88 (94.6)
Gastroesophageal junction	5 (5.2)	5 (5.4)
Histologic structure
Intestinal	17 (17.7)	22 (23.7)
Diffuse	60 (62.5)	53 (57.0)
Mixed/unknown	19 (19.8)	18 (19.4)
Previous gastrectomy	29 (30.2)	25 (26.9)
Previous neoadjuvant/adjuvant therapy	22 (22.9)	13 (14.0)
Measurable disease	55 (57.3)	54 (58.1)
No. of metastatic sites
≤2	74 (77.1)	79 (84.9)
≥3	22 (22.9)	14 (15.1)
Peritoneal metastases	63 (65.6)	56 (60.2)
Liver metastases	20 (20.8)	25 (26.9)
Ascites	37 (38.5)	33 (35.5)

Abbreviations: ECOG PS, Eastern Cooperative Oncology Group Performance Status; PBO+SOX, placebo plus S-1 and oxaliplatin; RAM+SOX, ramucirumab plus S-1 and oxaliplatin.

Most patients discontinued treatment in part A because of progressive disease (ramucirumab plus S-1 and oxaliplatin arm: 71 [74.0%] of 96 patients; placebo plus S-1 and oxaliplatin arm: 71 [76.3%] of 93 patients) (Figure 1). More patients discontinued treatment in part A because of an AE in the ramucirumab plus S-1 and oxaliplatin arm (11 [11.5%] of 96 patients) compared with the placebo plus S-1 and oxaliplatin arm (3 [3.2%] of 93 patients) (Figure 1). Overall, 118 patients (62.4%) received second-line paclitaxel plus ramucirumab treatment in part B (57 patients in the ramucirumab plus S-1 and oxaliplatin arm and 61 patients in the placebo plus S-1 and oxaliplatin arm) (Figure 1). (Patients in part B are referred to according to their treatment arm in part A: that is, ramucirumab plus S-1 and oxaliplatin or placebo plus S-1 and oxaliplatin.)

In part A, PFS was not prolonged in the ramucirumab plus S-1 and oxaliplatin arm compared with the placebo plus S-1 and oxaliplatin arm (Figure 2A). Median PFS was 6.34 (80% CI, 5.65-6.93) months in the ramucirumab plus S-1 and oxaliplatin arm and 6.74 (80% CI, 5.75-7.13) months in the placebo plus S-1 and oxaliplatin arm (HR, 1.07; 80% CI, 0.86-1.33; P = .70). Overall survival was similar between the 2 arms (Figure 2B). Median OS was 14.65 (80% CI, 12.39-15.67) months in the ramucirumab plus S-1 and oxaliplatin arm and 14.26 (80% CI, 13.83-17.31) months in the placebo plus S-1 and oxaliplatin arm (HR, 1.11; 80% CI, 0.89-1.40; P = .55). For PFS and OS, no notable improvements were observed in general in the ramucirumab plus S-1 and oxaliplatin arm compared with the placebo plus S-1 and oxaliplatin arm across prespecified subgroups (eFigure 2 in Supplement 2).

Figure 2.

Progression-Free Survival (PFS), Overall Survival (OS), and Second Disease Progression (PFS2) in Patients Receiving Ramucirumab Plus S-1 and Oxaliplatin (RAM+SOX) or Placebo Plus S-1 and Oxaliplatin (PBO+SOX) in Part A

Stratified log-rank test was used to analyze PFS, OS, and PFS2. Hazard ratios (HRs) were estimated with a stratified Cox proportional hazards regression model, and both were stratified by Eastern Cooperative Oncology Group Performance Status, region, and disease measurability. Median PFS was 6.34 (80% CI, 5.65-6.93) months in the RAM+SOX arm and 6.74 (80% CI, 5.75-7.13) months in the PBO+SOX arm (HR, 1.07; 80% CI, 0.86-1.33; P = .70). Median OS was 14.65 (80% CI, 12.39-15.67) months in the RAM+SOX arm and 14.26 (80% CI, 13.83-17.31) months in the PBO+SOX arm (HR, 1.11; 80% CI, 0.89-1.40; P = .55). Median PFS2 was 10.94 (80% CI, 9.63-12.52) months in the RAM+SOX arm and 11.99 (80% CI, 9.82-13.83) months in the PBO+SOX arm (HR, 1.11; 80% CI, 0.89-1.39; P = .55).

The overall response rate was 58% in the ramucirumab plus S-1 and oxaliplatin arm and 50% in the placebo plus S-1 and oxaliplatin arm (OR, 1.37; 80% CI, 0.84-2.24; P = .40). The disease control rate was 91% in the ramucirumab plus S-1 and oxaliplatin arm and 87% in the placebo plus S-1 and oxaliplatin arm (OR, 1.53; 80% CI, 0.68-3.43; P = .50). One patient (1.0%) in the ramucirumab plus S-1 and oxaliplatin arm and 3 patients (3.2%) in the placebo plus S-1 and oxaliplatin arm had a complete response.

The PFS2 was similar between the 2 arms (Figure 2C). Median PFS2 was 10.94 (80% CI, 9.63-12.52) months in the ramucirumab plus S-1 and oxaliplatin arm and 11.99 (80% CI, 9.82-13.83) months in the placebo plus S-1 and oxaliplatin arm (HR, 1.11; 80% CI, 0.89-1.39; P = .55).

Exposure

In part A, the median (range) relative dose intensity of ramucirumab was 91% (30%-107%) and the median (range) duration of ramucirumab treatment was 22.6 (2.0-99.4) weeks in the ramucirumab plus S-1 and oxaliplatin arm (Table 2). The median (range) relative dose intensities of S-1 (75% [15%-100%] vs 85% [34%-102%]) and oxaliplatin (80% [35%-111%] vs 86% [30%-106%]) were numerically lower in the ramucirumab plus S-1 and oxaliplatin arm compared with the placebo plus S-1 and oxaliplatin arm (Table 2). In addition, the median (range) durations of treatment of S-1 (21.1 [1.9-99.1] weeks vs 25.1 [1.6-102.1] weeks) and oxaliplatin (19.9 [3.0-90.0] weeks vs 22.3 [3.0-104.0] weeks) were numerically shorter in the ramucirumab plus S-1 and oxaliplatin arm compared with the placebo plus S-1 and oxaliplatin arm (Table 2). In part B, the relative dose intensities of ramucirumab were similarly high in the ramucirumab plus S-1 and oxaliplatin and placebo plus S-1 and oxaliplatin arms (99% vs 100%) (eTable 1 in Supplement 2).

Table 2.

Dose Intensity, Treatment Duration, and Cumulative Dose of Study Drugs in Part A

Study Drug	Variable	Median (Range)
		RAM+SOX Arm (n = 96)	PBO+SOX Arm (n = 93)
RAM/PBO	Relative dose intensity, %	91 (30-107)	94 (52-107)
	Duration, wk	22.6 (2.0-99.4)	24.0 (2.0-103.0)
	Cumulative dose per body weight, mg/kg	99.3 (3.2-490.4)	116.0 (8.0-504.8)
S-1	Relative dose intensity, %	75 (15-100)	85 (34-102)
	Duration, wk	21.1 (1.9-99.1)	25.1 (1.6-102.1)
	Cumulative dose, mg	8120.0 (600.0-44 290)	10 080.0 (420.0-52 140)
Oxaliplatin	Relative dose intensity, %	80 (35-111)	86 (30-106)
	Duration, wk	19.9 (3.0-90.0)	22.3 (3.0-104.0)
	Cumulative dose per body surface area, mg/m2	498.2 (91.9-1905.8)	583.1 (99.5-3239.8)

Abbreviations: PBO, placebo; PBO+SOX, placebo plus S-1 and oxaliplatin; RAM, ramucirumab; RAM+SOX, ramucirumab plus S-1 and oxaliplatin.

The geometric mean trough concentration levels of ramucirumab were as follows: 42.6 μg/mL on day 8 of cycle 1, 41.4 μg/mL on day 1 of cycle 2, 59.4 μg/mL on day 1 of cycle 3, 83.6 μg/mL on day 1 of cycle 5, and 94.6 μg/mL on day 1 of cycle 9 (eTable 2 in Supplement 2).

No new safety signals for ramucirumab were identified. In part A, the incidences of TEAEs (99% vs 100%) and treatment-related AEs (99% vs 99%) were similar in the ramucirumab plus S-1 and oxaliplatin and placebo plus S-1 and oxaliplatin arms. Serious AEs were reported by 28 patients (29.2%) in the ramucirumab plus S-1 and oxaliplatin arm and by 22 patients (23.7%) in the placebo plus S-1 and oxaliplatin arm. One treatment-related death occurred in the ramucirumab plus S-1 and oxaliplatin arm. The most frequently reported grade 3 or higher TEAEs in the ramucirumab plus S-1 and oxaliplatin arm were decreased neutrophil count (14 patients [14.6%]), hypertension (10 patients [10.4%]), and anemia (10 patients [10.4%]) (Table 3). The most frequently reported AEs of special interest in the ramucirumab plus S-1 and oxaliplatin arm were bleeding or hemorrhage events (36 patients [37.5%]), hypertension (28 patients [29.2%]), and proteinuria (24 patients [25.0%]) (Table 3).

Table 3.

Summary of Treatment-Emergent Adverse Events and Adverse Events of Special Interest in Part A

Grade	No. (%)
	RAM+SOX Arm (n = 96)		PBO+SOX Arm (n = 93)
	All	Grade ≥3	All	Grade ≥3
TEAEa
≥1 TEAE	95 (99.0)	66 (68.8)	93 (100.0)	55 (59.1)
Decreased neutrophil count	48 (50.0)	14 (14.6)	32 (34.4)	7 (7.5)
Hypertension	28 (29.2)	10 (10.4)	12 (12.9)	5 (5.4)
Anemia	24 (25.0)	10 (10.4)	22 (23.7)	11 (11.8)
Diarrhea	50 (52.1)	9 (9.4)	28 (30.1)	3 (3.2)
Decreased appetite	54 (56.3)	7 (7.3)	58 (62.4)	6 (6.5)
Decreased platelet count	33 (34.4)	6 (6.3)	28 (30.1)	3 (3.2)
Abdominal pain	18 (18.8)	3 (3.1)	24 (25.8)	5 (5.4)
AESIb
≥1 AESI	66 (68.8)	25 (26.0)	40 (43.0)	12 (12.9)
Bleeding/hemorrhage events	36 (37.5)	6 (6.3)	22 (23.7)	3 (3.2)
Hypertension	28 (29.2)	10 (10.4)	12 (12.9)	5 (5.4)
Proteinuria	24 (25.0)	4 (4.2)	14 (15.1)	1 (1.1)
GI hemorrhage events	5 (5.2)	3 (3.1)	5 (5.4)	1 (1.1)
Renal failure	5 (5.2)	2 (2.1)	2 (2.2)	0
GI perforation	3 (3.1)	3 (3.1)	2 (2.2)	1 (1.1)
Infusion-related reaction	2 (2.1)	1 (1.0)	3 (3.2)	0
Arterial thromboembolic events	1 (1.0)	1 (1.0)	0	0
Fistula	1 (1.0)	0	0	0
Venous thromboembolic events	0	0	3 (3.2)	2 (2.2)

Abbreviations: AESI, adverse event of special interest; GI, gastrointestinal; PBO+SOX, placebo plus S-1 and oxaliplatin; RAM+SOX, ramucirumab plus S-1 and oxaliplatin; TEAE, treatment-emergent adverse event.

Indicates TEAEs of grade 3 or higher occurring in at least 5% of patients in either arm.

Special interest category term.

In part B, in both arms, the most frequently reported grade 3 or higher TEAE was decreased neutrophil count, and the most frequently reported AE of special interest was bleeding or hemorrhage events (eTable 3 in Supplement 2).

Discussion

In this randomized phase 2 study (RAINSTORM), the addition of ramucirumab to first-line S-1 plus oxaliplatin treatment did not improve PFS, OS, or PFS2 in East Asian patients with metastatic gastric or gastroesophageal junction adenocarcinoma compared with S-1 plus oxaliplatin alone. Overall survival in both the ramucirumab plus S-1 and oxaliplatin and placebo plus S-1 and oxaliplatin arms was relatively longer compared with the OS found in the RAINFALL study,[13] which is consistent with previous studies conducted in East Asia.[9,19,20] Ramucirumab was associated with a manageable safety profile that was consistent with that reported in previous studies of ramucirumab.[7,8,13]

Two other studies[11,13] could not show a clinical benefit of adding ramucirumab to first-line chemotherapy for the treatment of metastatic gastric or gastroesophageal junction adenocarcinoma. In the randomized phase 3 RAINFALL study, conducted in 645 patients with metastatic gastric or gastroesophageal junction adenocarcinoma, the addition of ramucirumab to first-line fluorouracil-cisplatin or capecitabine-cisplatin chemotherapy resulted in a statistically significant improvement in investigator-assessed PFS (median PFS, 5.7 vs 5.4 months for chemotherapy alone; HR, 0.75; 95% CI, 0.61-0.94).[13] However, this improvement in PFS was not confirmed by central independent review of radiologic images (PFS HR, 0.961; 95% CI, 0.768-1.203).[13] Moreover, the addition of ramucirumab to first-line chemotherapy did not improve OS in the RAINFALL study.[13] In the randomized phase 2 JVBT (Ramucirumab Plus mFOLFOX as First-line Therapy in Patients With Advanced Esophageal, Gastroesophageal Junction, or Gastric Adenocarcinoma) study of first-line ramucirumab in 168 patients with advanced gastric cancer (including patients with esophageal adenocarcinoma), the addition of ramucirumab to mFOLFOX (oxaliplatin plus fluorouracil) did not improve PFS or OS (although in prespecified subgroup analyses, the PFS HR favored the ramucirumab arm in the gastric or gastroesophageal junction cancer subgroup but not in the esophageal cancer subgroup).[11]

A similar lack of effect on OS in patients with advanced gastric cancer in the first-line setting was observed for other VEGF-targeting agents such as bevacizumab, a monoclonal antibody against VEGF-A.[9,10] The chemotherapy backbone to the VEGF-targeting agent may be relevant to its activity. Paclitaxel is well known to show antiangiogenic activity[21]; therefore, its potential synergy with ramucirumab may have contributed to the improved outcomes observed in advanced gastric cancer in the second-line setting. Taken together, the RAINSTORM, RAINFALL,[13] and JVBT[11] studies, along with the bevacizumab studies,[9,10] suggest that adding an antiangiogenic agent to first-line fluoropyrimidine and platinum chemotherapy does not improve survival in patients with advanced gastric cancer. Further consideration of developing ramucirumab as a combination therapy in the first-line setting includes adding immunotherapeutic interventions.

A strength of this study was that its design incorporated a designated second-line treatment regimen, with the intention of having all patients receive second-line paclitaxel plus ramucirumab treatment. This study design might allow the assessment of the efficacy of continuous antiangiogenic therapy from first-line to second-line treatment. However, of the patients who received first-line treatment in part A, only approximately 60% of patients in both arms received second-line treatment with paclitaxel plus ramucirumab in part B. In total, 26 patients (16 in the ramucirumab plus S-1 and oxaliplatin arm and 10 in the placebo plus S-1 and oxaliplatin arm) received second-line treatment outside of part B because they did not meet the eligibility criteria for part B (4 patients received paclitaxel plus ramucirumab, 1 received ramucirumab monotherapy, and 4 received paclitaxel monotherapy). Furthermore, not all patients in part B received paclitaxel plus ramucirumab: 15 patients (13%) received either ramucirumab or paclitaxel monotherapy. Although the definition of PFS2 in this study took these factors into consideration, interpreting the efficacy of paclitaxel plus ramucirumab in part B of the study was difficult. However, the efficacy of second-line paclitaxel plus ramucirumab seemed to not be lessened because of first-line ramucirumab exposure.

Pharmacokinetic analysis showed that ramucirumab serum trough concentration levels in part A exceeded the biological target level (approximately 50 μg/mL[16]) after day 1 of cycle 3 and were generally consistent with those reported in the RAINFALL[13] and JVCX[16] studies, which also used a 3-week dose schedule. This schedule was chosen because higher levels of ramucirumab concentration have been associated with longer OS.[15] However, factors other than the ramucirumab concentration level, such as the chemotherapy partner to the VEGF-targeting agent, appear to be more influential in prolonging survival in patients who receive ramucirumab with a chemotherapy basis.

The safety profile of ramucirumab in the current study was manageable and consistent with previously published studies of ramucirumab in patients with metastatic gastric or gastroesophageal junction adenocarcinoma.[7,8,13] The study treatment was well tolerated. The incidences of neutropenia and thrombocytopenia in the RAINSTORM study were lower compared with the S-1 plus oxaliplatin arm of the G-SOX (S-1 Plus Oxaliplatin Compared With S-1 Plus Cisplatin as First-line Therapy for Advanced or Recurrent Gastric Cancer) study.[5] In the RAINSTORM study, the incidence of neutropenia of any grade was 50% and of grade 3 or higher was 15% in the ramucirumab plus S-1 and oxaliplatin arm and was 34% for any grade and 8% for grade 3 or higher in the placebo plus S-1 and oxaliplatin arm, compared with 68.9% for any grade and 19.5% for grade 3 or higher in the S-1 plus oxaliplatin arm of the G-SOX study.[5] The incidence of thrombocytopenia of any grade was 34% and of grade 3 or higher was 6% in the ramucirumab plus S-1 and oxaliplatin arm and was 30% for any grade and 3% for grade 3 or higher in the placebo plus S-1 and oxaliplatin arm, compared with 78.4% for any grade and 10.1% for grade 3 or higher in the S-1 plus oxaliplatin arm of the G-SOX study.[5] The AEs reported in part B of the study, in which patients received second-line paclitaxel plus ramucirumab, were consistent with those reported in the RAINBOW study of second-line paclitaxel plus ramucirumab.[8]

Limitations

This study has some limitations in addition to those discussed above. First, only approximately 60% of patients from part A received treatment in part B. Second, part B was not powered for the evaluation of PFS2.

Conclusions

Addition of ramucirumab to first-line S-1 plus oxaliplatin treatment did not improve PFS or OS compared with S-1 plus oxaliplatin alone in this phase 2 randomized clinical trial conducted in East Asian patients with metastatic gastric or gastroesophageal junction adenocarcinoma. The safety profile of ramucirumab observed in this study was consistent with that observed in previous studies of ramucirumab, and no new safety signals were identified. Although ramucirumab has been established as a treatment for advanced gastric cancer in the second-line setting, the RAINSTORM study, in combination with the RAINFALL and JVBT studies, suggest that the efficacy of ramucirumab as a second-line treatment may not be applicable to first-line treatment in combination with fluoropyrimidine and platinum chemotherapy.