H. Narahara 1 , W. Koizumi 2 , T. Hara 3 , A. Takagane 4 , T. Akiya 5 ,

1. Randomized phase III study of S-1 alone versus S-1 + cisplatin in the treatment of advanced gastric cancer (The SPIRITS trial) SPIRITS: S-1 plus cisplatin vs S-1 in RCT in the treatment of stomach cancer H. Narahara1, W. Koizumi2, T. Hara3, A. Takagane4, T. Akiya5, M. Takagi6, K. Miyashita7, T. Nishizaki8, O. Kobayashi9, S-1 Advanced Gastric Cancer (AGC) Clinical Trial Group; 1Osaka Medical Center for Cancer and CV Diseases, Osaka, JAPAN, 2Kitasato University East Hospital, Kanagawa, JAPAN, 3Kouseiren Takaoka Hospital, Toyama, JAPAN, 4Iwate Medical University, Iwate, JAPAN, 5Gunma Prefectural Cancer Center, Gunma, JAPAN, 6Shizuoka General Hospital, Shizuoka, JAPAN, 7National Hospital Organization Nagasaki Medical Center, Nagasaki, JAPAN, 8Matsuyama Red Cross Hospital, Ehime, JAPAN, 9Kanagawa Cancer Center, Kanagawa, JAPAN. SPIRITS

2. Background-1 • S-1 is : • an oral fluoropyrimidine widely used for AGC in Japan. • an oral formulation of Tegafur, CDHP, and Oxo at a molar ratio 1:0.4:1. • observed high RR and MST of 44-49 % and 207-250 days in two independent phase II trials1,2 Tegafur Anti-tumor activity CDHP Oxo 5-FU DPD OPRT GI toxicity F-b-Ala inhibit inhibit Neuro Toxicity Myelo toxicity 1: Y Sakata et al. Eur J Cancer 1998; 34: 1715-1720 2: W Koizumi et al. Oncology 2000; 58: 191-7

3. Background-2 • JCOG92051) • FP arm demonstrated significantly longer PFS than 5FU arm. (P<0.001) • There were no significant differences between the arms with respect to OS In Japan, recommended regimen for AGC was 5-FU alone 1): A. Ohtsu et al. J Clin Oncol 2003; 21:54-59

4. Background-3 5-FU Non-inferiority S-1 • JCOG9912 CPT-11+CDDP Boku et al. ASCO2007 abstract#: LBA4513

5. Background-4 • S-1+CDDP Phase I/II Study1) CDDP 60mg/m2 on Day 8 S-1 S-1 40-60mg BID for 3wks Day 15 Day 29 Day 36 Day 1 Day 8 Day 22 • Dosage of S-1 was based on patient’s body surface area (BSA) BSA < 1.25 : 40 mg BID 1.25 - < 1.50 : 50 mg BID 1.50 - < BSA : 60 mg BID 1: W Koizumi et al. Br J Cancer 2003; 89:2207-2212

6. Study Design • Central Randomization • (dynamic balancing) • Adjustment Factors: • Institute • PS • Unresectable vs Recurrent S-1 alone S-1: 40-60 mg BID for 28 days q6wks AGC No prior Chemo. R S-1 + CDDP S-1: 40-60 mg BID for 21 days q5wks CDDP: 60 mg/m2 iv on day 8

7. Endpoints • Primary Endpoint • Overall Survival • Estimated OS (S-1/S-1+CDDP) : 8/12 months • N=142 in each arm for 90% power to establish • superiority in OS (Two-sided log-rank a=0.05). • Follow up: 2 years • Secondary Endpoints • Progression Free Survival • Time to Treatment Failure • Overall Response • Safety 142 pts in each arm

8. Inclusion Criteria • Histologically confirmed gastric adenocarcinoma (unresectable/recurrent gastric cancer) • No prior chemotherapy • PS (ECOG scale) 0-2 • Age 20-74 • Expected survival > 3 months • Adequate organ function (bone marrow, liver, renal function) • Written informed consent

9. Patient Characteristics -1 • Randomized : 305 pts (S-1/S-1+CDDP : 152/153) • between Mar/2002 and Nov/2004 • FAS : 298 pts (S-1/S-1+CDDP : 150/148)

10. Patient Characteristics -2

11. Overall Survival Estimated probability (%) Log-rank p-value: 0.0366 HR: 0.774 [ 95% CI: 0.608 – 0.985] Median follow-up time (M): 34.6 11.0 13.0 Months

12. Progression-Free Survival Estimated probability (%) Log-rank p-value: <0.0001 HR:0.567 [ 95% CI: 0.437 – 0.734] 4.0 6.0 Months

13. Time to Treatment Failure Estimated probability (%) Log-rank p-value: 0.0089 HR:0.699 [ 95% CI: 0.536 – 0.912] 3.9 4.8 Months

14. Overall Response Fisher’s Exact Test p-value: 0.0018 • Criteria : RECIST (Extramural Review)

15. Adverse Drug Reactions-1 Criteria : NCI-CTC ver. 2.0

16. Adverse Drug Reactions-2 Criteria : NCI-CTC ver. 2.0 • No treatment-related death was observed

17. Phase III trials in AGC *TTP 3) Proc ASCO 2006; Vol 24, No. 18S: LBA4018 1) J Clin Oncol 2006; 24: 4991 – 4997 2) Proc ASCO 2006; Vol 24, No. 18S: LBA4017

18. Conclusions • The OS ofS-1+CDDP is superior to S-1 alone • The median survival time of S-1 was 11.0 M, moreover, that of S-1+CDDP was 13.0 M • S-1+CDDP is well tolerated and no treatment- related death was observed • S-1+CDDP regimen can be regarded as the first-line standard treatment for AGC

19. Future Perspectives In near future, the clinical characteristics of S-1+CDDP for AGC will become clear • In Red: FP vs S-1+CDDP (FLAGS: on-going) • In Yellow: S-1 vs S-1+CDDP (SPIRITS)

20. Acknowledgements Participating Inst. Kitasato University East Hospital Osaka Medical Center for Cancer and CV Diseases Kouseiren Takaoka Hospital Iwate Medical University Gunma Prefectural Cancer Center Shizuoka General Hospital National Hospital Organization Nagasaki Medical Center Matsuyama Red Cross Hospital Kanagawa Cancer Center (Department of Gastrointestinal Surgery) Hiroshima City Asa Hospital National Hospital Organization Kyushu Cancer Center Aso Iizuka Hospital Kanagawa Cancer Center(Department of Gastroenterology) Aichi Cancer Center Wakayama Medical University Saga Prefectural Hospital KOSEIKAN National Hospital Organization Tokyo Medical Center Osaka Saiseikai Nakatsu Hospital Aichi Cancer Center Aichi Hospital The Fraternity Memorial Hospital University of Tokai School of Medicine Gifu Municipal Hospital Toranomon Hospital Kumamoto Rosai Hospital Showa University Northern Yokohama Hospital Kobe City Medical Center General Hospital Kawaguchi Municipal Medical Center Nara Prefectural Nara Hospital Chikushi Hospital, Fukuoka University Kanto Medical Center, NTT EC Kyoto University Kinki University School of Medicine Yamagata Prefectural Central Hospital National Hospital Organization Fukuoka-Higashi Medical Center Kouri Hospital, Kansai Medical University Sendai Kousei Hospital Tokyo Women's Medical University Medical Center East Kokura Memorial Hospital Niigata City General Hospital This study was sponsored by TAIHO Pharmaceutical, co. ltd.