popescu razvan gastric cancer locally advanced

Treatment of operable gastric cancer

Razvan Popescu MD, MRCP(UK)

ESO Balkan Masterclass in Clinical Oncology 11.5.2011-‐ 15.5.2011 Dubrovnik, CroaKa

Gastric Cancer Incidence in Males GLOBOCAN 2008, Interna3onal Agency for Research on Cancer

0 3.2 6.9 11.6 21.9 63 Age-‐standardised incidence rates per 100,000

Cancer Incidence in Central and Eastern Europe GLOBOCAN 2008, Interna3onal Agency for Research on Cancer

•  Physical examinaKon, blood count and differenKal, liver and renal funcKon tests

•  Endoscopy / EUS •  CT scan of the thorax, abdomen and pelvis

•  Laparoscopy (+ peritoneal washings) –  +ve washings not independent prognosKc factor, conversion to –ve

washings up to 1/3 with preop chemotherapy (S. Lorenzen ASCO GI 2010)

•  PET scans –  can be negaKve, especially in paKents with mucinous tumours (up to 30%)

–  If posiKve can be used for early response assessment

Work-‐up of Gastric Cancer

Years after surgery

Gastric Cancer Survival

CADO,1985

0

50

100%

5 10 years

Stage III

Stage II

Stage I

Stage 0

21.9

47.6

79.2

91.6 82.0

66.9

36.4

14.7

Distal esophagus

Proximal stomach

Distal stomach

GE junction

EGJ Cancers and Gastric cancers are different enKKes !!

OGJ Cancers and Gastric cancers are different !!

•  Surgical resecKon is the only modality that is

potenKally curaKve, and is recommended for all non-‐

metastaKc cancers

•  The extent of opKmal regional lymphadenectomy is

sKll debated.

•  A minimum of 15 lymph-‐nodes should be recovered

(even if a formal D2 lymphadenectomy is not

performed)

Treatment of M0 Gastric Cancer

Dutch D1D2 surgical trial

•  996 eligible paKents randomized beteween 1989 and 1993 to D1 or D2 lymphadenectomy

•  771 paKents underwent assigned treatment, data reanalysed aaer 15 years

Outcome D1 D2 P

15-‐y survival 21% 29% 0.34

Gastric cancer death 48% 37% 0.01

Local recurrence 22% 12% -‐

OperaKve mortality 4% 10% 0.004

ComplicaKons 25% 43% 0.001

Strategies that increase cure rate in potenKally operable gastric cancer

•  Adjuvant chemotherapy •  Adjuvant Chemo-‐Radiotherapy

•  Peri-‐operaKve Chemotherapy

•  Pre-‐operaKve Chemotherapy, postoperaKve chemoradiotherapy

0.60 0.40 0.80 0.90 1.00 1.10 1.20 1.30 1.40 0.70 0.50 Surgery

alone better Any

chemotherapy better Hazard ratio

Overall effort HR: 0.82 (95% CI 0.76-0.91) P<0.0001

17 RCT 3838 pts

5 year survival: 55.3% vs. 49.6%

Benefit of adjuvant chemotherapy for resectable gastric cancer: a meta-analysis

JAMA. 2010 May 5;303(17):1729-3

•  Efficacy of treatment is unknown for the individual paKent

•  Results of individual trials discouraging – Some clearly underpowered to detect a significant survival difference. Other trials uKlized inferior surgical techniques.

•  Commencement of post-‐operaKve treatment may be delayed by slow recovery from surgery or peri-‐operaKve morbidity, problems with nutriKonal status

•  Treatment appears to be less well tolerated aaer major surgery

Challenges of adjuvant chemotherapy

POST-OP

PRE-OP

POST-OP

PRE-OP

Adjuvant chemotherapy: Percentage of Patients achieving adequate dose intensity

POST

PRE PRE

POST

SAKK TCF preop vs. postop Trial

•  4 cycles TCF planned either pre-‐ or postoperaKvely

•  Trial closed due to slow accrual (70 pats in 6 Y) •  PreoperaKve TCF given as planned in 74% of paKents, but only 34% postoperaKvely

•  SAE were more common in postoperaKve arm (23% vs. 11%)

SURGERY NO TREATMENT

RANDOMIZED N= 556 5-‐FU/FA x 1 (Mayo) STRATIFIED infusional 5-‐FU / 45 Gy

T 1-‐4 5-‐FU/FA x 1 (Mayo) NODES 0, 1-‐3, >3

McDonald JS et al. N Engl J Med 2001 Sep 6;345(10):725-‐30

Postoperative Chemoradiotherapy For Localised Gastric Cancer : INT-‐0116

•  Clear benefit in disease free and overall survival with median follow-‐up of 6 years. Risk reducKon of death by 24%.

•  Surgery: D2 resecKon less than 10%, 54 % of paKents fewer than 15 nodes (less than D1)

•  Planning of RadiaKon to be modified aaer central review in 35% of cases due to protocol deviaKons



•  Complex RT schedule with significant toxicity

•  SubopKmal chemotherapy schedule, role of the 2 flanking Mayo 5-‐FU/FA cycles unclear

Not an approach that has taken root in Europe

In the context of subopKmal surgery or if preoperaKve MDT is lacking an acceptable approach if good RT available



Impact of Extent of Surgery and PostoperaKve CRT on Recurrence Pamern

•  Leyden retrospecKve analysis of 2 Dutch trials : 91 paKents receiving postop CRT vs. Cohort from Dutch Gastric Cancer Trial (694) split by D1 vs. D2 resecKon

•  PaKents with D2 resecKon had as good an outcome as paKents receiving postop. CRT

•  Clear benefit for D1 resected paKents, R1 resecKons and high Maruyama Index of unresected disease (computed from data base of cases giving likelihood of involvement of unresected LN staKons)

JL Dikken, JCO May 10, 2010

Eligible patients: •  Adenocarcinoma of the stomach or lower third of the oesophagus (from 1999), suitable for curative resection •  Non-metastatic disease •  Stage II or greater

Chemotherapy (ECF): Epirubicin 50mg/m2, IV day 1 Cisplatin 60mg/m2, IV day 1 5-FU 200mg/m2/day, continuous infusion, days 1-21 (cycles repeated every 3 weeks)

Primary Overall survival

Secondary Progression-free survival Surgical resectability Quality of Life

Recruitment: July 1994-April 2002

MAGIC Trial Study entry and randomization

Pre-operative chemotherapy: ECFx3

Post-operative chemotherapy: ECFx3

Surgery

Surgery

S arm N=253

CSC arm N=250

3-6 weeks

6-12 weeks

Cunningham et al NEJM 2006

MAGIC Trial

CSC N=250

Commenced pre-operative chemotherapy N=237 (95%)

Completed pre-operative chemotherapy N=215 (86%)

Proceeded to surgery N=219 (88%)

Proceeded to surgery N=240(95%)

S N=253


MAGIC Trial Postoperative Morbidity/ Mortality

CSC S

Postoperative deaths 6% (14/219)

6% (15/24 0)

Postoperative complications 46% 46%

Median duration of post - operative hospital stay

13 days 13 days


MAGIC Trial Pathology Findings

•  Median maximum diameter of the resected tumor was smaller in the perioperaKve-‐chemotherapy group than in the surgery group (3 cm vs. 5 cm, P<0.001)

•  a greater proporKon of stage T1 and T2 tumors in the perioperaKve-‐chemotherapy group than in the surgery group (51.7 % vs. 36.8 %, P=0.002).

•  less advanced nodal disease (i.e., N0 or N1) in the perioperaKve-‐chemotherapy group than in the surgery group (84.4 % vs. 70.5 %, P=0.01)

MAGIC Trial Survival

Logrank p-value = 0.0001 Hazard Ratio = 0.66

(95% CI 0.53 - 0.81)

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

Months from randomisation 0 12 24 36 48 60 72

163 250 190 253

Events Total CSC S P

rogr

essi

on-fr

ee S

urvi

val r

ate Logrank p-value = 0.009

Hazard Ratio = 0.75 (95% CI 0.60 - 0.93)

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

Months from randomisation 0 12 24 36 48 60 72

149 250 170 253

Events Total

Sur

viva

l rat

e

CSC S

2 year survival

5 year survival

Median survival

CSC 50% 36% 24 mo

S 41% 23% 20 mo Benefit to CSC arm

9% 13% 4 mo

*Included relapse, PD and death from any cause.

PFS* Overall

  On multivariate analysis, treatment effect unchanged after adjustment for age, performance status, site of primary and gender

  Hazard ratio for death   Adjusted: 0.74 (95%CI:

0.59-0.93)   Unadjusted: 0.75


Cunningham D et al. N Engl J Med 2006;355:11-20

Tests for Heterogeneity of Treatment Effect According to the Baseline Characteristics of the Patients

FNLCC 94012-FFCD 9703 Trial

BOIGE et al ASCO 2007 *5-Fluorouracil 800 mg/m2 d1-5* + Cisplatin 100 mg/m2 day 1

FP (*) x 2/3 every 28 days

Resection

Within 4 weeks

4 - 6 weeks

Resection

4 – 6 weeks

FP x 3/4 or no treatment

Follow-up

Randomization N=224

CT + S S

Trial accrual 1995-2003

Median FU 5.7 yrs

2 year survival

5 year survival

Median survival

Periop CT 58% 38% 29 mo

Surgery 47% 24% 20 mo Benefit to CSC arm

10% 14% 9 mo

PFS* Overall

Median follow up: 5.7 years

___ S ___ CT + S

Logrank p value = 0.021 Hazard Ratio = 0.69 (95% CI 0.50-0.95)

  On multivariate analysis, treatment effect unchanged after adjustment for age, performance status, site of primary and gender

  Prognostic variables in Cox multivariate analysis:   Preoperative CT   Gastric location

FNLCC 94012-FFCD 9703 Trial

RFS OS

Pre-‐operaKve CT: the EORTC 40954 trial

144 paSents

resectable adenoca. of the stomach R

Surgery

PLF x 1 cycle

Surgery

PLF x 1 cycle

144 paSents randomized /360 in 4 years

Study prematurely closed because of poor accrual

Surgery

Restaging If NO PD/tox/WHO 2

N= 72

N= 72

Neoadjuvant Arm

Surgery arm

p

R0 resecSon 59 (81.9%) 48 (66.7%) 0.036

N0 node 27 (38.6%) 13 (19.1%) 0.018


DFS OS


ResecKon Rates

MAGIC (n=503)

FFCD (n=224)

EORTC (n = 114)

S Chemo+S S Chemo+S S Chemo+S

96% 92% 99% 96% 94% 96%

Survival Hazard RaKos

MAGIC (n=503)

FFCD (n=224)

EORTC (n = 114)

S Chemo+S S Chemo+S S Chemo+S

0.75 0.69 0.84

Summary pre-‐ /perioperaKve Chemotherapy

•  All trials suggest a down sizing and down staging of gastric cancers, no relevant risk of progression whilst on chemotherapy, no increased complicaKons perioperaKvely and improved PFS and OS

•  No standard chemotherapy regimen – choose best advanced chemotherapy available

Pre-‐ / peri-‐operaKve Chemotherapy, postoperaKve chemoradiotherapy

•  RaKonale: sKll high rates of local failure, may be improved by postoperaKve RT, now combined to modern preoperaKve chemotherapy (mostly ECF or modificaKons thereof)

•  Trials ongoing •  Not standard, may be appropriate in paKents with expected poor surgical results (e.g. insufficient LN dissecKon, high number /raKo of involved / resected LN)

How to improve benefit of systemic treatment

•  Improve regimens

•  Tailor treatment

Role of PET in idenKfying paKents who may benefit from neo-‐adjuvant chemotherapy

PET -‐Responders

PET –NON Responders

PET -‐Responders

PET –NON Responders

How to improve benefit of systemic treatment

•  Improve regimens

•  Tailor treatment

•  Establish working mulKdisciplinary teams that meet regularly and mandate that oncological treatment should be first discussed in an MDT

Summary

•  Systemic treatment improves outcome of operable gastric cancer in all seqngs

•  PreoperaKve approaches are preferred – Bemer delivery of treatment – Monitoring of response – Downstaging and downsizing of tumor

•  Ongoing research regarding opKmal regimen and tailoring of treatment

•  MDT essenKal in improving management

popescu razvan gastric cancer locally advanced

Health & Medicine