The rise of a fallen star : Re-emerging role of RT for pancreatic cancers

1. The rise of a fallen star : Re-emerging role
of RT for pancreatic cancers
Dr Bala Vellayappan MBBS, FRANZCR, MCI
Consultant Radiation Oncologist
National University Cancer Institute Singapore
Assistant Professor, NUS
3rd Meeting of the Federation of
Asian Organizations for Radiation
Oncology (FARO)
Bali
8th September 2018

2. Content for today
• Part 1 : Background and definition
• Part 2 : Rise and fall (and hopefully rise again)
• Part 3: Evidence and technique for SBRT

3. Pancreatic cancer is deadly
• Fourth leading cause of mortality
• Median survival times
• 20 % : Resectable : 24 - 36m
• Borderline resectable : 15 – 18 m
• 40% : Locally advanced/unresectable : 10 – 15m
• 40% : Metastatic : 6- 12 m

4. Surgery remains the gold standard for cure
Ryan NEJM 2014 p1039
Toesca IJORBP 2018
En bloc resection : distal stomach, duodenum, Head
of pancreas, distal bile duct,GB, prox jejunum

5. Content for today
• Part 1 : Background and definition
• Part 2 : Rise and fall (and hopefully rise again)
• Part 3: Evidence and technique for SBRT

6. Evidence Base for the use of RT
PLUS MINUS
Adjuvant / (neoadjuvant) GITSG 1
RTOG 9704 2
GERCOR Phase II 3
PREOPANC 4
ESPAC-1…. 5
Definitive ECOG 4201 (CRT vs chemo) 7
GITSG (CRT vs chemo) 8
FFCD/SFRO 2000-01 (CRT vs chemo) 9
ECOG 8282 (CRT vs chemo) 10
LAP07 11
EORTC 6
1. Kalser 1985
2. Regine 2008
3. Van Laethem 2010
4. Van Tienhoven 2018 ASCO LBA
5. Neoptolemos 2001
6. Klinkenbijl 1999
7. Loehrer 2011
8. JNCI 1988
9. Chauffert 2008
10. Klaassen 1985
11. Hammel 2016

7. Evidence Base for the use of RT
PLUS MINUS
Adjuvant / (neoadjuvant) GITSG 1
RTOG 9704 2
GERCOR Phase II 3
PREOPANC 4
ESPAC-1…. 5
Definitive ECOG 4201 (CRT vs chemo) 7
GITSG (CRT vs chemo) 8
FFCD/SFRO 2000-01 (CRT vs chemo) 9
ECOG 8282 (CRT vs chemo) 10
LAP07 11
EORTC 6
1. Kalser 1985
2. Regine 2008
3. Van Laethem 2010
4. Van Tienhoven 2018 ASCO LBA
5. Neoptolemos 2001
6. Klinkenbijl 1999
7. Loehrer 2011
8. JNCI 1988
9. Chauffert 2008
10. Klaassen 1985
11. Hammel 2016

8. Chemo-RT in the adjuvant setting
Trial N Regimen LF OS Significance Comments
GITSG
21
22
40Gy (split) +5FU
Obs
NR
NR
20m
11m
P= 0.03
Poor accural
EORTC
60
54
40Gy + 5FU
Obs
NR
NR
17.1m
12.6m P=0.099
Mixed
periampullary and
pancreatic
ESPAC-1
73
75
72
69
40Gy + 5FU (split)
5FU alone
chemoRTchemo
obs
NR 14m
22m
20m
17m
Authors
concluded adj
chemo beneficial
(even in R1), adj
chemoRT
detrimental
RTOG 9704
230
221
5FU-> CRT +5 FU->5FU
Gem->CRT + 5FU->gem
30%
25%
17.1m
20.5m
P=0.12
MVA showed gem
regimen has a
trend towards
improved OS
GERCOR
(Phase II) 45
45
Gem alone
Gem + gem-RT
24%
11%
24m
24m
NS Less local
recurrence with
CRT

9. RT kills patients…

10. Should we believe ESPAC-1?
• Designed to be a 2 x 2 RCT
• Obs (surgery alone)
• Adj ChemoRT only
• Adj Chemo only
• Adj chemoRT + chemo
• But due to poor accrual, there was an additional 2 randomisations (+-
adj chemoRT, +- adj chemo)

11. Things to note…
• Treatment not delivered in 30% of patients
• Significant protocol violations in all arms
• Cross-over allowed
• Patients could be given “background” therapy, which could be chemo or RT
• Poor accrual allowed physicians allowed to choose which arm patient goes to
• Why ESPAC-1 may have been negative
• 2/3 had local relapse
• Inadequate radiotherapy dose? [40Gy in 2 Gy fraction, split course]
• Poor quality RT?
• Editorial suggested that the detriment in survival may be related to RT-related morbidity (because
the observation arm had better survival than the chemoRT arm!!)
• Addition of a local treatment was not associated with improvements in survival –
patients need to live long enough for this translation.

12. Geography Current standard Comments
USA R0 : chemo alone or chemoRT +
chemo (ACR)
R1 or factors for local recurrence :
chemoRT + chemo
RTOG 0848 will clarify the use of
chemo--chemoRT
Europe R0 and R1 : chemo alone ESPAC 2,3,4
PRODIGE 24/CCTG PA.6 (median OS
54 vs 35 m) – 40% had R1 resection
Singapore (NCIS) R0 : adjuvant chemo alone
R1 : chemoRT + chemo
Sandwich regimen. Gem x 1 –50.4
to 59.4Gy IMRT + Xeloda - -
adjuvant Gem

13. Borderline resectable
Toesca et al IJORBP 2018

14. Rationale for neoadjuvant therapy
• Positive resection margin status portends a poorer prognosis1
• Aim of neoadj therapy : downstage and sterilise the tumour boundaries in
contact with vessels (tumour-vessel interface) to facilitate a margin
negative resection (R0)
• Higher chance that patients will complete all planned systemic treatment
• Less hypoxia
• Less complications2
• Smaller volume (usually no ENI)
• Allows for weeding out patients with poor disease biology from undergoing
unnecessary surgery
• Under debate : NA chemo alone or NA chemo + RT or NA chemo—
chemoRT or NA chemo—SBRT3
1.Ghaneh. Ann Surg. 2017
2.Ishikawa. Arch Surg 1991 p885
3.Toesca. IJROBP 2018 p1155

15. GemG
RT: 36 Gy in 15 fractions of 2.4 Gy
Immediate surgery
N=127
Preop. radiochemotherapy
N=119
p value
Resection rate 91/127 (72%) 72/119 (60%) .065
R0 resection rate PP 28/91 (31%) 45/72 (63%) <.001
Serious Adverse Events 49 (39%) 55 (46%) .28
Preoperative radiochemotherapy versus immediate surgery for
resectable and borderline resectable pancreatic cancer (PREOPANC)
Van Tienhoven
LBA4002 ASCO 2018

16. Overall survival (ITT)
Median survival:
13.7 vs 17.1 months
HR 0.74
p=0.074
Preliminary: 149/176 events
Van Tienhoven
LBA4002 ASCO 2018

17. Disease-free survival
Median DFS:
7.9 vs 9.9 months
HR 0.71
p=0.023
Van Tienhoven
LBA4002 ASCO 2018

18. Metastases and local recurrence (ITT)
HR 0.55
P = 0.002
HR 0.71
P = 0.013
Van Tienhoven
LBA4002 ASCO 2018

19. Unresectable (Locally advanced)

20. Evidence Base for the use of RT
PLUS MINUS
Adjuvant / (neoadjuvant) GITSG 1
RTOG 9704 2
GERCOR Phase II 3
PREOPANC 4
ESPAC-1…. 5
Definitive ECOG 4201 (CRT vs chemo) 7
GITSG (CRT vs chemo) 8
FFCD/SFRO 2000-01 (CRT vs chemo) 9
ECOG 8282 (CRT vs chemo) 10
1. Kalser 1985
2. Regine 2008
3. Van Laethem 2010
4. Van Tienhoven 2018 ASCO LBA
5. Neoptolemos 2001
6. Klinkenbijl 1999
7. Loehrer 2011
8. JNCI 1988
9. Chauffert 2008
10. Klaassen 1985
11. Hammel 2016
LAP07 11
EORTC 6

21. CRT in the definitive setting
Trial N Chemo LF OS Significance Comments
GITSG
CRT (40Gy)
Chemo alone
22
21
5FU + SMF
SMF
45%
48%
9.7m
7.4m
P<0.02
Trial closed due
to poor accural
ECOG
CRT (40Gy)
Chemo alone
47
44
5FU
5FU
32%
32%
8.3m
8.2m
NS
FFCD/SFRO
CRT (60Gy)
Chemo alone
59
60
5FU/cis
Gem
NR 8.6m
13m
P=0.03 in
favour of
chemo
RT 60Gy +
5FU/cis
possibly too
toxic
ECOG 4201
CRT (50.4Gy)
Chemo alone
34
37
Gem
Gem
12%
30%
11m
9.2m
P=0.017

22. LAP 07 – should we interpret
this as a plus or a minus for RT?

23. LAP 07
• P : Locally advanced pancreatic cancer (i.e unresectable). N= 449
• I : 2 x 2 factorial : First randomisation gem +- Erlotinib
• After 4 months of systemic therapy, 2nd randomisation chemoRT
(54Gy + Cape) vs chemo alone (N=269)
• O: Primary outcome : OS [HR 1.03 95% CI 0.79 – 1.34]
• Secondary outcome : PFS, toxicity

24. Better local control and longer time-off
systemic therapy!
• Median PFS : 8.4m vs 9.9m (HR 0.78, 95% CI 0.61 - 1.01 , p=0.06)
• Loco-regional progression was less frequent with chemoRT (32 vs 46%, P=0.04)
• Median delay to re-starting chemo was longer with chemoRT. (6.1 vs 3.7m,
P=0.02)
• 4% managed to undergo curative resection

25. With improved systemic therapy, durable local control
becomes important
Patients need to have longer
survival to reap the benefit of RT.
Perhaps with multiagent systemic
therapy (i.e FOLFIRINOX) 
improved OS RT has a bigger
role to play to optimise local
control
Groot. Ann Surg 2018 p936

26. CRT vs chemo (Definitive)
Trial N Chemo LF OS Significance Comments
GITSG
CRT (54 Gy)
Chemo alone
22
21
5FU + SMF
SMF
45%
48%
9.7m
7.4m
P<0.02
Trial closed due to
poor accural
ECOG
CRT (40 Gy)
Chemo alone
47
44
5FU
5FU
32%
32%
8.3m
8.2m
NS
FFCD/SFRO
CRT (60 Gy)
Chemo alone
59
60
5FU/cis
Gem
8.6m
13m
P=0.03 in favour of
chemo
RT 60Gy + 5FU/cis
possibly too toxic
ECOG 4201
CRT (50.4 Gy)
Chemo alone
34
37
Gem
Gem
12%
30%
11m
9.2m
P=0.017
LAP07
Chemo—CRT(54 Gy)
Chemo—more chemo
133
136
Gem-5FU
Gem
32%
46%
15.2m
16.5m
NS Negative for primary
endpoint

27. Is there a benefit with dose escalation?
Unresectable Ca
Pancreas
Induction
chemotherapy :
FOLFIRINOX or
gem-based
Cape+RT 50.4Gy/28#
Dose escalation given to patients
where tumour was >1cm from GI
structures (BED >70Gy)
Krishnan. IJROBP 2016 p755

28. • Higher dose(BED)was the only predictor of improved OS on multivariate analysis.
• No additional toxicity was observed in the high-dose group
BED>70Gy had a superior OS (17.8 vs 15.0
months, P=.03)
Improved locoregional RFS (10.2 v 6.2m P=0.05)

29. Can we identify patients who are prone to fail
locally?
• Based on a previous autopsy series, 30% of patients died due to a
locally destructive CaP, and the remainder due to widely metastatic
disease .
• Intact SMAD4 expression was highly correlated to the locally
destructive pattern (loss of SMAD4 = distant mets)
Locabuzio-Donahue JCO2009 p 1806

30. Treatment paradigm for borderline
resectable/unresectable Ca P
Induction chemotherapy
FOLFIRINOX for good PS
Gem-based for elderly, lower KPS
4-6 months for BRPC, >6m for unresectable PC
Re-evaluate with imaging
Stable disease or responders – proceed to RT
PD or distant mets – 2nd line chemo
ChemoRT (Dose-escalated IMRT with xeloda)
Or SBRT 30-40Gy in 5 fractions

31. Content for today
• Part 1 : Background and definition
• Part 2 : Rise and fall (and hopefully rise again)
• Part 3: Evidence and technique for SBRT

32. SBRT for pancreatic cancer
• Personal opinion : Survival for pancreatic cancer remains short. A 6-
week course of RT takes up ~10% of their remaining life span!
• Hypofractionated RT/SBRT adds to QUANTITY of life outside of the
hospital
• 5-6 weeks of RT may also delay time off effective multi-agent systemic
therapy ?higher incidence of distant relapse

33. The WHY and HOW
Rationale for SBRT
• Ability to selectively escalate the
dose to the tumour-vessel
interface facilitates a margin-
negative resection and
decreases the risk of local
relapse
Successful SBRT
• Treats tumour only, no ENI!
• Accounts and reduces organ
motion
• Achieves sharp dose fall-off from
surrounding normal tissue
• Requires image guidance to
maintain precision
Reduce volume of normal tissue
irradiated and thereby reduce
toxicity

34. Early experience from Stanford
• 77 patients, unresectable
• 25Gy single fraction (Cyberknife)
• 96% received gem-based chemotherapy
• 1 year freedom from local progression 84%
• 9% G3 acute toxicity,
• 25% G2+ late toxicity
Chang et al Cancer 2009; 115: 665

35. Multi-fraction SBRT appears to be safer and
more effective
Pollom IJROBP 2014 p 918
Sutera Front Oncol 2017 p 272
25Gy x 1 fractions
VS
33 Gy in 5 fractions

36. Duodenal morbidity is the biggest obstacle for
pancreatic SBRT
• Majority of Ca P occur in the head, as duodenum surrounds the
tumour
• Pooled data now provide us with prescriptive limits for the duodenum
Goldsmith. Sem in Rad Onc 2016

37. Geus, Cancer 2017 p 4158
Unadjusted mean survival : Chemo 9.9m < chemo +EBRT
10.9m < chemo+IMRT 12m < CHEMO + SBRT 13.9m

38. Ongoing RCTs – watch this space
A Randomized Phase III Study Evaluating
Modified FOLFIRINOX (mFFX) With or Without
Stereotactic Body Radiotherapy (SBRT) in the
Treatment of Locally Advanced Pancreatic
Cancer (PanCRS) – PI : Dan Chang
(ALLIANCE) : preoperative extended
chemotherapy vs. chemotherapy plus
hypofractionated radiation therapy for
borderline resectable adenocarcinoma of the
head of the pancreas – PI : Matt Katz
RT : SBRT 33Gy/5# , or HIGRT
25Gy/5# (?3DCRT)

39. Ongoing RCTs – watch this space

40. How to do it?
• Wait 1 week post EUS-guided fiducial
implantation
• Supine, arms up, patient fasted 3-4 hours
• Small amount of water 50 ml (to visualise
duodenum)
• Abdominal compressor, other motion
management strategies
• Multiphasic helical CT (pancreatic and PV
phase)
• Helical contrast scan (for target) and 4DCT
(for motion and planning)
• IGRT

41. IGRT : Bone matching alone is not adequate
Jayachandran IJROBP 2010 p 603

42. • Alignment of seeds and stents
compared separately to planning
CT images
• Aligning to stents lead to a
7.7mm displacement from seeds
• Stent alignment lead to reduced
coverage of PTV
• Fiduicial>stents>bone matching
Pepin PRO 2015

43. Planning case
study
• 60/F, ECOG 1
• Unresectable Ca P
(involvement of SMA
and celiac) T4N0M0
• Confirmed with EUS-FNA
: adenoca
• Metal biliary stent
inserted for obstructive
jaundice

44. 6 FOLFIRINOX 
Response on CT and
Ca19-9.
No distant mets, still
unresectable
Symptomatic with celiac
plexus pain

45. Opted to have SBRT : 40Gy in 5 fractions to TVI (red colourwash) , 30Gy in 5 fractions to PTV
(green colourwash)

46. Suggested constraints for 5 # regimen
Organ at Risk Constraint
Liver V12 <50%
Kidney (combined) V12<75% 200cc <17.5Gy
Cord 25Gy max V20<1cc
Stomach V33<1cc, V20<3cc, V15<9cc,
V12<50%
Duodenum V33<1cc, V20<3cc, V15<9cc,
V12<50%
Dmax <35
Bowel V20<5cc

47. Conclusion
• Majority of patients still die from their cancer.
• 1/3 of patients treated with chemotherapy alone have a local
recurrence – morbid!
• We need better ”systemic therapy” and ”local therapy”
• Effective yet convenient treatment is preferred
• Multi-fraction (3-5) SBRT has lower toxicity than single fraction
• Neoadjuvant RT is useful to increase the R0 resection rate

48. Bala_vellayappan@nuhs.edu.sg