Transcript ASCO_2008_files/Ilson Discussant GEJ ASCo 2008 oral

Esophageal and Gastric Cancer: How
do we sort out the treatment options?
David H. Ilson, MD, PhD
Gastrointestinal Oncology Service
Memorial Sloan-Kettering Cancer Center
Disclosure

Research Funding
– Sanofi-aventis
– Genentech
– Bristol-Myers Squibb/Imclone
Esophageal and Gastric Carcinoma
US Incidence in 2008

37,970 new cases
– Gastric:
21,500 (57%)
– Esophagus:
16,470 (43%)

Decline in Gastric Cancer , Esophageal Squamous
Cancer Incidence

Increase in Adenocarcinoma of the esophagus, GE JX,
cardia

Esophageal cancer more virulent:
– 87% fatality rate
– Gastric: 52%
Jemal et al, CA 58: 71-96; 2008
Oral Presentations: Esophageal and
Gastric Cancer, Phase III

Adjuvant Therapy

Igaki (Abs 4510): Esophageal Cancer
– Pre versus Post op chemo in squamous cancer

Kang (Abs LBA 4511): Gastric Cancer
– Adding IP and IV cisplatin to MMC / oral doxifluridine after
resection

Advanced Disease

Ridwelski (Abs 4512): Gastric Cancer
– Docetaxel + cisplatin versus 5-FU + cisplatin
Esophageal Cancer
Trials of Adjuvant Therapy

Preoperative  surgery

Preop Chemo  Surgery

Preop Concurrent RT + Chemo + / Surgery
A randomized trial of postoperative
adjuvant chemotherapy with
cisplatin and 5-fluorouracil versus
neoadjuvant chemotherapy for
clinical Stage II/III squamous cell
carcinoma of the thoracic
esophagus (JCOG 9907)
H. Igaki, N. Ando, H. Kato, M. Shinoda, H. Shimizu, T. Nakamura,
S. Ozawa, H. Yabusaki, N. Aoyama, A. Kurita, H. Fukuda
Japan Esophageal Oncology Group (JEOG) of
Japan Clinical Oncology Group (JCOG), Japan
6
Prior JCOG Trials in Esophageal
Squamous Cancer

Similar stage patients, similar sample size and design

JCOG 8806: Surgery vs post op vindesine + cisplatin,
205 pts
– Negative Trial
– OS 5 year 45% for surgery, 48% for post op chemo (NS)

JCOG 9204: Surgery vs post op 5-FU + cisplatin, 242 pts
– Primary: DFS 5 year
– Negative Trial
– DFS 45% for surgery, 55% for chemo (NS)
– OS in node + patients: 38%  52% post op chemo
(P = 0.041)

Unplanned subset analysis
Ando J Card Thor S 114: 205; 1997 Ando J Clin Oncol 24: 4592;2003
Scheme
Randomization
Balanced with minimization by
institution, cN0 / cN1
Post-op CTx
(standard arm A)
Pre-op CTx
(test arm B)
Surgery
2 x FP
2 x FP
Surgery
In pN0,no CTx
FP: cisplatin + 5FU
5-FU
800mg/m2 d1-5 ci
cisplatin 80mg/m2 d1 div
Surgery
-Transthoracic esophagectomy
with lymphadenectomy (>D2)
8
Igaki et al, Abstract 4510

Primary Endpoint: increase PFS by 10-13% for
preoperative arm

Adequately powered trial
– Protocol excluded pathologic N0 pts on post op arm (38 pts)
from chemo
– Based on analysis of prior post op trial: no benefit in N0

Arms balanced by preclinical stage

Therapy tolerable and feasible

Primary endpoint not reached

Median PFS: 2 year (post op) versus 3 year (pre op, NS)

OS 5 year 38% post op  60% pre op (HR 0.64, p = 0.014)
2nd Interim Analysis
• To decide early publication or not @ Mar. 2007, all 330 pts
Overall survival (OS)
Progression-free survival (PFS)
1.0
1.0
0.9
0.9
Pre
0.8
Median PFS=3.0y
0.7
0.6
0.5
0.5
0.4
0.4
Post
0.2
Post
0.3
0.2
Median PFS=2.0y
0.1
5yOS=60.1%
0.7
0.6
0.3
Pre
0.8
5yOS=38.4%
0.1
0.0
0.0
0
1
2
3
4
5
6
7
0
1
Years after randomization
Unadjusted one-sided stratified
logrank P =
0.0444
Hazard ratio =
0.76
> 0.0254 (alpha)
(94.91%CI: 0.56–1.04)
2
3
4
5
6
7
Years after randomization
Unadjusted two-sided logrank
P = 0.013
Hazard ratio by Cox model
=
0.64
(95%CI: 0.45–0.91, p=0.014)
DSMC recommended the early publication
10
Igaki et al, Abstract 4510




A large number of post op pts did not receive chemo
– Preop: 17 pts (10%) did not receive chemo
– Post op: 85 pts (51%) did not receive chemo
 Includes 38 path N0 pts excluded per protocol (23%)
 Another 28% failed to receive post op chemo
Trial did not reproduce results from post op arm of prior
study
– 61% OS prior post op trial
– 38% OS current post op trial
Different populations, pathologic staging on the prior
trial compared to clinical staging on the current trial
Adequacy of clinical staging: EUS and PET scan were
not performed
– ? Treatment arms truly balanced by stage
Subgroup Analyses: updated OS
Data in Nov 2007
Clinical Stage
Stage II
1.0
Pre
(n=80)
5yOS=69.7%
0.9
0.8
0.9
0.7
0.6
0.6
0.5
0.5
0.4
0.4
Post
0.1
(n=84)
5yOS=52.1%
Post
0.3
(n=78)
5yOS=49.3%
0.2
Pre
0.8
0.7
0.3
Stage III (non-T4)
1.0
0.0
(n=88)
5yOS=36.5%
0.2
0.1
0.0
0
1
2
3
4
5
6
7
8
0
1
2
Years after randomization
Hazard ratio by Cox model
=
0.48
(95%CI: 0.28–0.83)
Two-sided P =
0.0088
3
4
5
6
7
8
Years after randomization
Hazard ratio by Cox model
=
0.94
(95%CI: 0.61–1.46)
Two-sided P =
0.79
Pre-op chemotherapy may be more beneficial in Stage II
12
Igaki et al, Abstract 4510

Current Preop Trial: only Stage II patients benefit

Prior post op trial benefit limited to path N + pts

Dose this trial prove that pre op chemo is superior to
post op chemo in esophageal squamous cancer?

No

Unplanned subset analyses generate hypotheses and
not conclusions

Large number of pts did not receive chemotherapy

Adequacy and accuracy of pre treatment clinical
staging
Esophageal Cancer: Preop Chemotherapy

Negative Trials

U.S. INT 113
100
– 440 pts
80
– Adeno 54%, Squamous
60
46%
40
– 3 pre, 3 post op cycles
of 5-FU + Cisplatin
– No improvement in
20
0
0.5
1
1.5
2
disease free or overall
survival
Kelsen et al, NEJM 339: 1979; 1998
2.5
3
3.5
4
4.5
5
Esophageal Cancer: Preop Chemotherapy

Positive trials

U.K. MRC OEO-2
– 802 pts
– Adeno 66%, Squamous 31%
– 2 preop cycles of 5-FU + Cisplatin

9% increase in 2 year OS

Decreased to 6% increase in 5
years OS (17%23%)

10% operative mortality
MRC Lancet 359: 1727; 2002
Cunningham NEJM 355: 11; 2006
Meta Analysis Preop Chemo in
Esophageal Cancer: Thirion (ASCO 2007)

Squamous and adeno
– 9 trials OS: 2102 pts

Overall survival improvement for preop chemo
(HR 0.87, p = 0.0033)
– 4% improvement in OS at 5 yrs
– Squamous 4%
– Adeno 7%
Preop Chemo in Esophageal Squamous
Cancer

Marginal benefit from larger trials and
meta analysis

Operative mortality risk = survival
benefit

Data do no support this approach as the
optimal standard of care in squamous
cancer
Preop Chemo in Esophageal Squamous
Cancer

Primary chemo + radiotherapy: U.S. standard of care for
locally advanced disease
– Curative potential without surgery (RTOG)
– Responding patients (FFCD)
 Absence of a clear survival benefit for addition of
surgery
– Surgery for biopsy positive / non responding patients after
therapy
– Surgery for younger, fitter patients

Early stage disease:
– Surgery alone or Chemoradiotherapy alone
Gastric Cancer:
Adjuvant Therapy Standards of Care

Post op Chemo oral S-1 (Japan)
– 10% improved 5 yr OS
– 1000 pts, D2 resection

Post op Chemo 5-FU / LV + radiotherapy (U.S.)
– 10% improvement in 5 yr OS
– Pts with less than a D1-D2 resection

Pre and post op Chemo ECF (U.K.)
– 13% improvement in 5 yr OS
– without radiotherapy
Sakuramoto NEJM 357: 1810;2007 Macdonald NEJM 345: 725; 2001 Cunningham NEJM
355: 11; 2006
Rationale: Intraperitoneal Chemotherapy

Peritoneal recurrence in up to 40-50% of
patients after gastric resection

IP chemo increases exposure to chemo
agents by 10-100 + fold

No confirmed phase III trials of IP
therapy
Postoperative adjuvant chemotherapy for grossly
serosa-positive advanced gastric cancer:
A randomized phase III trial of intraperitoneal
cisplatin and early mitomycin-C plus long-term
doxifluridine plus cisplatin (iceMFP) versus
mitomycin-C plus short-term doxifluridine (Mf)
(AMC 0101) (NCT00296322)
Yoon-Koo Kang, Heung-Moon Chang, Dae Young Zang,
Jae-Lyun Lee, Tae Won Kim, Dae Hyun Yang, Se Jin Jang,
Jeong Hwan Yook, Sung Tae Oh, Byung Sik Kim
Asan Medical Center, Seoul, Hallym University Hospital, Anyang, Korea
ASCO 2008 Abs 4531: Adjuvant MMC + FU + / IV Cisplatin after D2 resection (AMC 0201)

MMC x 1, Doxifluridine orally x 3 mos
– vs

MD + IV Cisplatin x 6, Doxifluridine x 12 mos

MD: 67% 3 yr relapse free survival

MDP: 65% 3 yr relapse free survival

No improvement with addition of IV cisplatin or extended
therapy

Current Trial (AMC 0101) MMC + FU  + IV cisplatin +
extended doxifluridine
– + IP cisplatin x 1 intra op
– Mitomycin administered earlier on
– Serosa positive gastric cancer
Treatment Schema
Grossly Serosa(+), Non-Metastatic Gastric Cancer
RANDOMIZATION
iceMFP arm
Intraperitoneal CDDP
MMC
At surgery
Stratified by center, stage
Mf arm
100 mg for 2h
before closure
3 - 6 weeks
after surgery
15 mg/m2 iv D1
4 weeks later
CDDP
DFUR
CDDP
DFUR
CDDP
DFUR
CDDP
DFUR
CDDP
DFUR
CDDP
DFUR
DFUR
DFUR
DFUR
DFUR
DFUR
DFUR
DFUR
Stage I,
IV(M1)
20 mg/m2 iv
MMC
Protocol off
DFUR
DFUR
DFUR
DFUR 460 – 600 mg/m2 po daily
Cisplatin 60 mg/m2 iv D1 every 4 weeks
Recurrence Free Survival
Recurrence free proportion
1.00
0.75
0.50
0.25
N
Event
3yRFSR
5yRFSR
iceMFP
263
103
60.2%
50.5%
Mf
258
126
50.0%
43.8%
HR 0.695 [ 95% C.I.: 0.536 - 0.902 ]
P = 0.006 by log-rank test
0
0
12
24
36
48
months after randomization
60
72
Overall Survival
Surviving proportion
1.00
0.75
0.50
0.25
N
Event
3yOSR
5yOSR
iceMFP
263
82
71.2%
56.2%
Mf
258
103
59.6%
47.0%
HR 0.710 [ 95% C.I.: 0.531-0.950 ]
P = 0.02 by log-rank test
0
0
12
24
36
48
months after randomization
60
72
Recurrences
P=0.02
Intraperitoneal Cisplatin, Gastric Cancer

Well designed and conducted phase III trial

Adequately powered

Ineligible patients common
– Randomization was intraoperative prior to final
pathology reading
– Appropriate exclusion of metastatic disease, positive
margins

Arms balanced, therapy tolerable, no increased
operative complications
Intraperitoneal Cisplatin, Gastric Cancer

Achieved primary endpoint of improved
relapse free survival:
– 3 yr RFS 50%  60%

Reduced peritoneal and distant recurrence

Suggests a potential benefit for IP cisplatin,
early start of mitomycin
Intraperitoneal Cisplatin, Gastric Cancer

Four variables confound conclusion that IP
chemo is superior in serosa + gastric cancer
– Trial also included early MMC, extending
doxifluridine, adding IV cisplatin

Companion AMC 0201 trial
– No benefit for extended chemo or addition of
cisplatin

Confirmatory trial in which only variable is IP
chemotherapy
Esophago-Gastric Cancer: Metastatic
Disease
•
CIV 5-FU + cisplatin
• 4-5 day infusion
• RR 20-30%
• Med S 8-9 months
•
Adding a third drug:
• Epirubicin (ECF), protracted 5-FU CIV: RR 40-45%, Med S 9
mos
• 25-40% have locally advanced, non metastatic disease
• Docetaxel (DCF): RR 36%, Med S 9 mos
• 10% increment in response rate
• 1-2 month increment in survival
•
Capecitabine = 5-FU, Oxaliplatin = Cisplatin
DCF vs CF: Toxicity
Grade 3/4 Toxicity
DCF
CF
Stomatitis
21%
27%
Diarrhea
19%
8%
Nausea/vomiting
14%
17%
Neutropenia
82%
57%
Neutropenic fever
29%
12%
Taken off toxicity
of AE
Toxic deaths
49%
37%
3.6%
5.4%
Van Cutsem J Clin Oncol 2006
4512
Docetaxel–cisplatin (DC) versus 5-fluorouracil–
leucovorin–cisplatin (FLC) as first-line treatment for
locally advanced or metastatic gastric cancer:
Preliminary results of a Phase III study
K Ridwelski, J Fahlke, C Schmidt, E Kettner, U Keilholz,
D Quietzsch, M Assmann, M Stauch, K Zierau, H Lippert
Objectives
Primary objective
Time to tumor progression (TTP)
increase from 40% to 60% in patients without
progression at 6 months
Secondary objectives
Toxicity
Overall response rate (ORR)
Overall survival (OS)
Study design
Chemotherapynaïve patients with
metastatic or
locally advanced
(stage III-IV)
gastric cancer
R
A
N
D
O
M
I
Z
A
T
I
O
N
DC
Docetaxel 75 mg/m2 iv over 1 h, day 1
Cisplatin 75 mg/m2 iv over 1 h, day 1
Repeated q3w for up to 6 cycles
FLC
5-FU 2000 mg/m2 iv over 24 h
Leucovorin 500 mg/m2 iv over 2 h
Both on days 1, 8, 15, 22, 29, 36
+
Cisplatin 50 mg/m2 iv over 1 h, days 1, 15, 29
Repeated q7w for up to 4 cycles
(cisplatin omitted in Cycle 4)
273 patients randomized between October 2001 and September 2006 from 21 centers in Germany Data available for 270 patients
Results
DC
FLC
Median TTP, months
n=133
5.8
n=137
6.6
Median OS, months
8.2
9.6
1-year survival data
28 %
34 %
TTP
P=0.468
OS
P=0.616
Median follow up:
8.2 months
9.5 months
DC, docetaxel–cisplatin; FLC, 5-FU–leucovorin–cisplatin; OS, overall survival; TTP, time to progression
Response to treatment
DC
FLC
n=133
0
n=137
4
Partial response (PR)
32
29
Stable disease
59
59
Progressive disease
26
25
Not evaluable
16
20
Response, n
Complete response (CR)
ORR (CR + PR), % (95% CI)
p-value
24.1 (0.17-0.32)
24.1 (0.17-0.32)
p = 0.244
Confirmed response per RECIST
DC, docetaxel–cisplatin; FLC, 5-FU–leucovorin–cisplatin; ORR, overall response rate
Safety
Grade 3/4 adverse event (NCI-CTC)
patients
Hematologic ( % )
Leukopenia
Neutropenia
Febrile Neutropenia
Anemia
Non-hematologic ( % )
Gastrointestinal
Nausea
Vomiting
Pain
Infection
DC, docetaxel–cisplatin; FLC, 5-FU–leucovorin–cisplatin
DC
n=127
FLC
n=134
45.2
8.2
46.2
5.5
12.2
0.0
15.1
10.4
18.5
17.3
8.7
13.4
5.6
29.2
22.2
15.5
12.6
5.2
DC vs CF

DC is not superior to CF, DC = CF
– Contrasts the V325 study: D + CF superior to CF

CF well tolerated on this dose and schedule

Use colorectal cancer like schedule for
cisplatin or oxaliplatin + FU
– Weekly or two weekly infusional FU
– Single digit toxicities
Colorectal Scheduling: Gastric Cancer, Weekly
infusional 5-FU + every 2 week Oxaliplatin or
Cisplatin
Grade 3/4
Toxicity
Stomatitis
FLO
FLP
3%
3%
Diarrhea
3%
6%
Nausea/vomiting
5%
9%
Neutropenia
5%
9%
Neuropathy
14%
2%
Off therapy due
to toxicity or AE
17%
23%
Al Batran J Clin Oncol 26: 1435; 2008
DC vs CF

DC = CF, failed to achieve primary endpoint of
superiority

Contrast to V325 study, three drug DCF superior to CF
– Toxicity of DCF limits its general use

If DCF is used:
– Lower starting doses
– Consider colorectal like scheduling of drugs

Marginal differences in outcome for three drugs:
Doublets better platform for future study
– Adding targeted agents
– Perioperative therapy or added to radiotherapy
Gastric Cancer Chemotherapy: What
regimen to use?
Oxaliplatin
Cape
XP
FLO
FUFIRI
DCF
ECF
EOX or
EOF
ECX or
EOX
Pts
489
513
160
109
170
221
126
%RR
44%
45%
41%
34%
32%
36%
45%
TTP, mos 6.7
6.5
5.6
5.5
5.0
5.6
7.4
OS, mos
10.4
10.5
--
9.0
9.2
8.9
10.9