Should we use chemotherapy or chemoradiation for borderline resectable pancreatic cancer?

1
Should we use chemotherapy or chemoradiation for
borderline resectable pancreatic cancer?
Jordan D. Berlin, M.D. Ingram Professor of
Medicine/ Clinical Research Co-director, GI
Oncology Director, Phase I Research Vanderbilt-Ing
ram Cancer Center
2
Disclosures

• Advisory Boards here and there in last year
• Genentech/Roche
• Sanofi-Aventis
• Abbott
• Amgen
• Astra Zeneca
• Clavis/Clovis
• Infiinity
• Otsuka
• Arno
• BMS
• Synta

• Research Support
• Amgen, Genentech/Roche, Lilly/Imclone,
  OSI/Astellas, Bayer, Pfizer, Novartis
• Honorarium
• Lilly
• Added this week
• Symphogen
• Lilly

3
Stage at DiagnosisWe catch disease late
Unstaged
15
Localized disease
Metastatic disease
8
52
Locally advanced disease
26
Ries et al (eds). SEER Cancer Statistics Review,
1975-2001.
4
5-Year Survival by StageStill Poor at All Stages
We have not really gone far since this data so
there is room for improvement
Ries et al (eds). SEER Cancer Statistics Review,
1975-2001.
5
TNM staging vs Practical Staging

• In pancreatic cancer, we have not traditionally
  made our decisions based purely on TNM staging
• Instead, we had 3 stages
• Localized, resectable
• Localized, unresectable
• Metastatic
• Now. We have a new group in town
• Borderline resectable
• These are patients whose disease appears able to
  be removed surgically
• However, there is a high probability of margin
  positivity

6
Defining Borderline Resectable Pancreas Cancer is
Complex

• Surgically resectable is in the eye of the
  beholder
• The lines of what can be removed have become
  increasingly blurred with
• New techniques for vascular reconstruction
• Varying surgical skills
• Variable radiologic studies/techniques
• Pathologic margins are not always
  reported/identified in the same way
• Retroperitoneal margin/SMA margin

7
MD Anderson Definition of Borderline Resectable

• Includes (not all at once)
• Abutment of the SMA involving lt 180 of the
  circumference of the artery
• Encasement of a short segment of the hepatic
  artery, without evidence of tumor extension to
  the celiac axis, that is amenable to resection
  and reconstruction
• Short-segment occlusion of the SMV, PV, or SMPV
  confluence with a suitable option for vascular
  reconstruction
• Does not address colon and/or mesocolon invasion

Varadhachary, et al Annals of Surgical Oncology
13 1035-1046
8
AHPBA Consensus Conference Refined the MDACC
Criteria

• Includes
• No distant metastases.
• Venous involvement of the SMV/portal vein
  demonstrating tumor abutment with or without
  impingement and narrowing of the lumen,
  encasement of the SMV/portal vein but without
  encasement of the nearby arteries,
• or short segment venous occlusion resulting from
  either tumor thrombus or encasement but with
  suitable vessel proximal and distal to the area
  of vessel involvement, allowing for safe
  resection and reconstruction

Callery, et al Ann Surg Oncol (2009) 1617271733
9
AHPBA Consensus Statement II

• Continued
• Gastroduodenal artery encasement up to the
  hepatic artery with either short segment
  encasement or direct abutment of the hepatic
  artery, without extension to the celiac axis
• Tumor abutment of the SMA not to exceed 180 of
  the circumference of the vessel wall.
• Note Method of assessment Multidetector CT
  scan, 2 phases, with 3-dimensional reconstruction
• NCCN guidelines use this definition

Callery, et al Ann Surg Oncol (2009) 1617271733
10
Separate European Guidelines Dont Exist

• No definition of borderline resectable in ESMO
  guidelines, but treatment consideration is as
  follows
• Patients who present borderline resectable
  disease may benefit from preoperative therapy
  (chemoradiation or induction chemotherapy
  followed by chemoradiation) in order to increase
  the rate of R0 resections.
• ESDO met this AM to develop expert opinion
• Next slide

Cascinu, et al Ann Oncol (2010) 21 (suppl 5)
v55-v58
11
ESDO Expert Opinion meeting Identified Key
Weaknesses in the Guidelines

• AHPBA, NCCN and MD Anderson all use very similar
  definitions of borderline resectable, but
• We are unclear if we have truly separated
  unresectable from borderline resectable
• Terms about the radiographic findings including
  impingement, abutment, involvement and encasement
  while used commonly may not be as clearly defined
  as they could be
• I will add that we have not tested to see if
  there is consistency between radiologists in
  defining these findings

12
Differing Criteria for Borderline Resectable may
Produce Different Results

• Recent study of FDR gemcitabine capecitabine as
  neoadjuvant therapy (no XRT)
• Local criteria used 33 borderline resectable
  patients and 10 unresectable patients
• NCCN criteria used 18 borderline resectable, 25
  unresectable
• By local criteria, 15 BR went to surgery, 13 R0
  resections and 1 of 2 UR who went to surgery had
  R0 resection
• By NCCN criteria, 11 BR went to surgery, 9 R0
  resections and 5 of 6 UR who went to surgery had
  R0 resection

Lee J-L, et al Surgery 2012, epub ahead of print,
June 6
13
3 Principle Goals of Neoadjuvant Therapy

• Response
• This may not be RECIST response
• Needs to sterilize the margins
• Needs to shrink away from the vessels if possible
• Margin free resection
• All data suggests that margin resections result
  in poorer survival outcomes
• Not interfering with surgical outcome
• Treatment should not cause increased
  morbidity/increased post-operative complications
• Treatment should not cause fibrosis/scarring that
  make the operation more difficult

14
Combined Analysis of Published Data Shows Low
Response Rates
CR PR SD PD
All patients (n 330) 1.8 18.8 59.2 18.9
Resectable (n 196) 0.8 9.5 73.9 17
Borderline/unresectable (n 134) 4 31.8 40.9 21.8
54.2 of all patients underwent resection 65.8
of resectable patients unwderwent
resection 80.6 of these were R0 31.6 of
borderline/unesectable patients underwent
resection 62.2 of these were R0 All patients
received chemo, 85 had chemoxrt
Mura Assifi, et al Surgery 150466-73, 2011
15
Combined Analysis Shows We Can Achieve RO
Resection

• Suggests that neoadjuvant therapy leads to high
  R0 resection rate
• These studies had differing definitions of
  resectable, borderline and unresectable
• Intriguingly, borderline and unresectable
  patients who had resection had the same survival
  (22.3months) as resectable patients (23 months)
• Does this suggest our definitions of borderline
  resectable are just bad on these studies?
• Did not differentiate chemo from chemoradiation

16
Does chemoradiation have a higher response rate
than chemo alone?

• Very little evidence of this
• Even in the combined analysis, the definitions of
  response varied over the years
• Primary pancreatic cancers
• Appear less responsive than metastases (this is
  different from most other tumor types)
• Are difficult to measure even with high quality
  scans

17
E4201 Locally Advanced pancreatic Cancer Trial
Schema
RANDOMIZE
ARM A CONSOLIDATION GEMCITABINE 1000mg/M2 Once
weekly x 3 weeks Followed by 1 week rest x 5
cycles 1 cycle 4 weeks
ARM A INDUCTION GEMCITABINE 1000mg/M2 Once
weekly x 6 weeks
1 week rest

• Stratify
• PS (0 vs 1)
• Weight loss
• ( gt10 vs lt10)

ARM B INDUCTION GEMCITABINE 600 mg/M2 Once
weekly x 6 weeks CONCURRENT RT 180 cGy/day 5
days week x 6 weeks Total dose 50.40 Gy
ARM B CONSOLIDATION GEMCITABINE 1000mg/M2 Once
weekly x 3 weeks Followed by 1 week rest x 5
cycles 1 cycle 4 weeks
4 weeks rest
18
E4201 Response is the Same for Chemo and ChemoXRT
GEM alone N 35 GEM plus XRT N 34
Partial Resp. 5 6
Stable Disease 35 68
Progression 16 6
Inevaluable 46 21
Clinical progression without confirmation
scans or scans performed outside of scheduled
times
19
E4201 But ChemoXRT has More Stable Disease
GEM alone N 35 GEM plus XRT N 34
Partial Resp. 5 6
Stable Disease 35 68
Progression 16 6
Inevaluable 46 21
Clinical progression without confirmation
scans or scans performed outside of scheduled
times
20
E4201 ChemoXRT May Decrease Incidence of Local
Relapse
GEM alone GEM plus XRT
Local 41 23
Distant 14 23
Local and Distant 5 9
Not documented 41 44
Clinical progression without confirmation
scans or scans performed outside of scheduled
times
21
E4201 Conclusions

• Despite no improvement in response or PFS
• This study appeared to show a modest survival
  benefit (time was equal to the time patients
  spent getting radiated)
• While response was not higher for XRT, local
  relapse appeared modestly less likely when XRT
  was used
• Does not give conclusive evidence of anything as
  it was underpowered due to poor accrual

22
Can Chemotherapy Before ChemoXRT Provide Better
Outcomes

• 70 patients with borderline (n 24), or
  unresectable (n 46) disease treated with
  chemoXRT
• Two strategies
• ChemoXRT with 50.4Gy (53 unresectable
  pre-treatment)
• Chemo (gem based) followed by ChemoXRT if no PD
  after chemo (83 unresectable pre-treatment)
• 20 in both strategies had resection
• The patients who underwent chemo followed by
  chemoradiation had an improved OS (18.7 vs 12.4
  months, p 0.02) compared to chemoXRT alone

Arvold ND, et al Cancer 20121183026-35
23
Is FOLFIRINOX the neoadjuvant regimen of choice?
R A N D O M I Z E
Folfirinox
Metastatic pancreatic cancer
Gemcitabine

• Stratification
• center
• performance status 0 versus 1
• location of the tumor head versus other location
  of the primary

24
FOLFIRINOX Response Rate is High
Folfirinox N171 Gemcitabine N171 p
Complete response 0.6 0
Partial response 31 9.4 0.0001
CR/PR 95 CI 24.7-39.1 5.9-15.4
Stable disease 38.6 41.5
Disease control CRPRSD 70.2 50.9 0.0003
Progression 15.2 34.5
Not assessed 14.6 14.6
Median durationof response 5.9 mo. 4 mo. ns
25
Overall Survival
Median follow up 26.6 months 95 CI 20.5
44.9
Folfirinox N171 Gemcitabine N171 p HR
Median survival CI 95 11.1 mo. 9 - 13.1 6.8 mo. 5.5 - 7.6 lt0.0001 0.57
1-yr. survival 48.4 20.6
18-mo. survival 18.6 6
26
FOFLIRINOX Conclusions

• It is a more effective regimen in every parameter
  (RR, PFS, OS) and significantly so
• It is logical (synergy between oxali and each of
  the other two drugs)
• It is more toxic than gemcitabine
• Key point all patients were metastatic and it is
  not clear the response rate would be as high with
  local disease only

27
FOLFIRINOX Neoadjuvant Analysis There is Concern
it is not Enough

• Retrospective analysis of 18 patients treated
  with neoadjuvant FOLFIRINOX
• Borderline and unresectable patients
• 14 borderline resectable by AHPBA guidelines
• 4 resectable after treatment by imaging, 2 of
  whom had R0, 1 had R1 and 1 was unresectable
• 3 of 3 unresectable patients deemed resectable
  after treatment had R0 resection

28
It is Not Clear RECIST is the Parameter we Should
Use to Say a Regimen Helps

• MDACC Retrospective Analysis
• 122 of 129 patients with borderline resectable
  disease
• RECIST 1.1 was used
• PR 15 patients (12) All resected
• SD 84 patients (69) 70 (84) resected
• PD 23 patients (19) 0 resected
• Changed to resectable 1 (0.8) resected
• 85 patients underwent resection
• 95 had R0 resection
• Median OS was 33 months (12 months for
  unresected)
• RECIST response did not predict survival

Katz MHG, et al Cancer, epub 2012
29
Pathologic Response is a Criteria?

• Retrospective review was designed to compare 32
  patients with borderline resectable disease
  treated with chemoXRT then surgery to 104
  patients with resectable disease taken to surgery
• 18/32 patients had pathologic response
• Degree of pathological responses was graded based
  on the proportion of fibrosis and necrosis
  replacing primary tumor involvement and was
  estimated as percentages of the total tumor
  volume.
• Pathologic Response correlated with T stage
• Not sure if this is a legitimate criteria as it
  is subjective and the disease has fibrosis and
  necrosis without treatment

Kang CM, et al J Gastrointest Surg (2012)
16509517
30
Alliance A021101 Protocol

• Pre-Registration allows for
• Biliary decompression
• Central review of staging scans (restaging also
  reviewed centrally)

31
Alliance A021101 Protocol

• Endpoints
• Primary
• Estimate the 1-year overall survival (OS) rate
• Secondary
• To estimate the rate of treatment-related
  toxicity during preoperative therapy.
• To estimate the R0 resection rate following
  preoperative therapy.
• To estimate the rate of radiographic and
  histopathologic response to preoperative therapy.
• To estimate the time to locoregional and distant
  recurrence following completion of treatment.

32
Other benefits of Alliance A021101 Protocol

• There is currently no reported, completed
  multicenter clinical trial focusing on borderline
  resectable disease, so this may
• Establish a baseline for future study
• Establish parameters for assessing margins
• Establish standards for borderline resectable
  disease
• By radiographic criteria
• Establish surgical standardization

33
Borderline Resectable Conclusions

• We need to establish more standards for this
  category
• Pathology
• Surgery
• Radiology
• Treatment choice
• These definitions need to truly separate
  borderline unresectable from truly unresectable
  patients
• Standard of care is not clearly defined
• It is clear we need more and better studies.