1. Ph positive ALL; Is transplant still necessary. Adele K
Ph positive ALL; Is transplant still necessary? Adele K. Fielding MB BS, PhD
Chair UK National Cancer Research Institute
Adult ALL Subgroup

2. Outline of talk
Why we use alloHSCT in Ph+ ALL – what is the evidence it is of value?
Why is it so difficult to draw firm conclusions about the role of alloHSCT?
Where does non-myeloablative alloHSCT fit in?
Now we have TKIs, do we still need transplant
Illustrate various points above with some data from UKALL12/ECOG2993
Conclusions about current role of TKI and alloHSCT

3. Ph+ ALL - the scope of the problem
Long been concluded that patients with t(9;22) have a poor outcome with conventional therapy
Considered “Very high risk” by all clinical studies
Typically ‘assigned’ to alloHSCT

4. Ph+ ALL - the scope of the problem
Karyotype
*% Adults
**% Children
t(4;11) 5 1
t(9;22) 21 2
t(12;21)
<1 19
High hyperdiploidy 9 32
Poor risk
Good risk
*Moorman et al Blood 2007
**Hann et al BJH 2002

5. Age in ALL; tolerance of therapy ?
Group Induction Resistance to
Death % induction Rx “Children”
AIEOP
COALL
DFCI
EORTCCCLG
SJ CRH 13B
“Adults”
GMALL05/
GOELAM
HyperCVAD 5 3
JALSG
LALA
After Pui and Evans, NEJM 2006

6. Myeloablative allogeneic HSCT
The most active anti-ALL therapy currently available, but also the most toxic
For whom?
Age
Status of disease
At what level of relapse-risk
Convincing data
Useful data
Conflicting data
What is the best method ?
Conditioning regimen
Stem cell source
Unrelated vs. sibling donor
Role of T cell depletion
GVHD prophylaxis/ KGF e.g. palifermin
Post-BMT interventions e.g. DLI, TKI
Scant data
There are few published data on optimal conditioning regimen but TBI is a key component. An IBMTR study suggested a benefit to the combination of etoposide with TBI The role of T cell depletion is also unclear – practice in this area tends to be institutionally or nationally-based. Unrelated donors are increasingly used as a source of stem cells and data suggest little difference in the outcome, given good matching . With increasing use of unrelated donors, the true role of alloHSCT may become almost impossible to evaluate formally.

7. GVL effect The most active anti-ALL therapy currently available,
but also the most toxic
Fielding et al Blood 2009 UKALL12/ECOG2993

8. OS > 600 patients after relapse
At what level of relapse-risk?
100 75
Fielding et al 2007 UKALL12/ECOG2993
Percent surviving 50 7% 25 1 2 3 4 5
Time (years)
Other large relapse studies:
Tavernier et al LALA94 DFS 5 yr 12%
Oriol et al PETHEMA OS 5yr 10%

9. Myeloablative alloHSCT A ‘transplanters’ perspective?
“That which doesn’t kill us makes us stronger”
Nietzsche
“If it doesn’t feel bad, it can’t be doing you any good”
Alice Fielding (my mum)

10. Sibling Myeloablative allo HSCT
Pre-TKI era
*prospective studies

11. UD donor Myeloablative allo HSCT; preTKI era
Retrospective studies, not all patients Ph+
Mixed ages
Some patients beyond CR1
Equivalent outcome between unrelated and sibling donor:
Cornillissen, 2001 Blood
Dahlke, 2006 BMT
Kiehl, 2004 J Clin Oncol

12. UD donor Myeloablative allo HSCT OS UKALL12/ECOG2993
Pre-TKI
100 75
PERCENT 50
44 Sib
36 MUD 25
19 chemo 1 2 3 4 5
At risk:
TIME IN YEARS
Sib 45 35 29 25 19 18
MUD 31 23 12 12 11 11
Chemo 82 43 23 19 15 12
Fielding et al Blood 2009

13. Difficulties in studying the alloHSCT question
1. Trial Design
Equipoise problems in designing RCTs with myeloablative alloHSCT as an arm
Impossibility of “donor vs. no donor”analysis in modern transplant studies where UD
are commonly available

14. Difficulties in studying the alloHSCT question
2. Selection bias
“Transplant only” retrospective studies do not include a denominator. Hence, benefit can only apply to those selected for reporting
Those might be only a tiny fraction of the at-risk population & are unlikely to be representative

15. Difficulties in studying the alloHSCT question
3. Immortal time bias
A patient receiving a transplant within a prospective study is guaranteed to have entered CR and survived, disease free to the time of transplant
A simple analysis by therapy received during trial doesn’t represent the reality for a future new patient accurately

16. Immortal time bias: an example
Fielding et al, Blood 2009 UKALL12/ECOG2993 – pre TKI era

17. Immortal time bias: an example
Fielding et al Blood 2009
UKALL12/ECOG2993

18. Myeloablative Allo in Ph ALL
Weight of evidence has been interpreted
In favour of myeloablative allo HSCT
Those patients who
Achieve CR
Have a donor
Are fit enough to have a transplant
…Receive the transplant
Will have a better outcome than those who don’t

19. Non-myeloablative alloHSCT

20. Non-myeloablative alloHSCT
Likely to be
much more regimen-dependent,
more dependent on disease burden at allo
require more focused post-transplant monitoring (+ intervention ?) than myeloablative

21. TKI in Ph ALL 1. Is there a higher CR rate with no excess cost ?
2. Do TKI facilitate alloBMT?
3. Is there a survival advantage for TKI-containing regimens?
4. Does this advantage occur in the absence of alloHSCT
For purpose of this talk TKI = largely Imatinib
Data on dasatinib during induction are much fewer/shorter follow up

22. TKI in Ph+ ALL Cautionary note
Hu et al. Nat Genet. 2004
Requirement of Src kinases for BCR-ABL-induced B-ALL
but not CML.
Pfeifer et al Blood 2007
BCR-ABL KD mutation in 40% at presentation
Soverini et al Hematolgica 2011
2-4 clones of 200 for each patient harboured
pre-exisiting KD mutations

23. Induction therapy Ph pos ALL:Imatinib
Imatinib added to induction/consol’n.
N CR(%) CR1 SCT (%)
Thomas
Yanda
Wassmann
De Labarthe
Ribera
Bassan
Fielding
Chalandon ongoing /100
Older
Ottman N/A
Vignetti N/A
Children only sibs
Schultz 92

24. Induction therapy Ph pos ALL:Imatinib
Imatinib added to induction/consol’n.
N OS time
Thomas m
Yanda m
Wassmann /43 24m
De Labarthe m
Ribera yr
Bassan yr
Fielding yr
Chalandon ongoing yr
Older
Ottman m
Vignetti m
Children
Schultz (EFS) 3yr

25. Induction therapy Ph pos ALL:Dasatinib
Dasatinib added to induction/consol’n.
N CR(%) CR1 SCT (%) combo
Ravandi hyperCVAD
Foa N/A Steroid
Rousselot N/A EWALL elderly backbone

26. Induction therapy Ph pos ALL:Dasatinib
Dasatinib added to induction/consol’n.
N OS time
Ravandi m
Foa m
Rousselot 71 median 21.7m

27. TKI without transplant in Ph+ ALL
Older patients
Patients unfit for alloHSCT
Those fit, but without any donor
Children, in whom there is reluctance
to use UDs

28. TKI without transplant in Ph ALL
NoBMT
BMT
JALSG: Yanada et al, JCO 2006

29. Children with Ph+ ALL Risk:benefit ratio of Rx is unfavourable for UD
Arico et al, NEJM 2009

30. Children with Ph+ ALL; COG study
N=25 chemo+ im 87.7%
N=11 UD HSCT 71.6%
N=25 sibHSCT 56.6%
Schultz et al, 2009 J Clin Onc

31. UKALLXII/ECOG2993 Ph+: up to 65y
INDUCTION
Phase 1
Phase 2
Early Imatinib N=89
response
determination
Late Imatinib (as a consolidation) N=86
Allogeneic HSCT
(Etoposide/TBI)
up to 55y.
no donor/unfit
for allo
Autologous HSCT
(Etoposide/TBI)
recommend
All patients IMATINIB 600mg od

32. Patient characteristics
* Pre-imatinib results: Fielding et al, Blood 2009, “imatinib” paper Fielding et al Blood 2014

33. Response to Induction

34. Flow Chart of Post-Induction Therapy
Induction death/No CR 14 (8%)
Protocol deviation induction 24 (14%)
Chemo/imatinib
39/137 eligible (28%)
39/175 total (22%)
Total alive/CR
137 (77%)
Myeloablative allo HSCT
82/137 eligible (60%)
82/175 total (46%)
43 sibling donor
33 Matched unrelated donor
3 cord blood
2 Mismatched unrelated donor
1 Haploidentical donor
“Non-protocol” RIC alloHSCT
11/137 eligible (8%)
11/175 total (6%)
6 sibling donor
5 Matched unrelated donor

35. Transplant details Protocol BMT rate Pre-imatinib 28% Imatinib 46%
Reach 84 day in CR 78% 66% .009
Median days diag to BMT 135d 160d <.0001
% elig pt receiving BMT 71% 55% <.0001
Both imatinib, overall better BMT practice contributed to the
improved alloHSCT rate after imatinib

36. Overall outcomes Imatinib vs.Pre-Imatinib
Median FU Pre-imatinib 10 years
Median FU Imatinib 3 years

37. Imatinib vs. pre-ImatinibOS
Imatinib vs. pre-Imatinib 1 2 3 4 5 6 7 8 9 10 25 50 75
100
19%
32%
At risk:
266
144 81 67 57 54 49 46 41 40 35
Imatinib
175
115 69 53 31 17
Pre-imatinib
imatinib
Pre-imatinib versus imatinib, p=0.0003
Percent surviving
Time (years)
Pre-imatinib

38. Imatinib vs. Pre-Imatinib
Relapse risk
Imatinib vs. Pre-Imatinib
100
64% Imatinib 75
% relapsing 50
47% Pre-imatinib 25 1 2 3
Time in years
2P = 0·0001

39. OS within imatinib cohort from diagnosis, by treatment in CR1
100 75
Protocol alloHSCT
50%
Percent surviving 50
Non-protocol RIC alloHSCT
39% 25
19%
No alloHSCT 1 2 3 4
At risk:
Time (years)
Protocol alloHSCT 76 57 45 42 31
Non-protocol RIC alloHSCT 11 8 6 5 3
No alloHSCT 38 20 9 7 6

40. Relapse risk within imatinib cohort by treatment in CR1
100
73% No BMT 75
% relapsing
49% Non Protocol
BMT 50 25
25% Prot BMT 1 2 3
Time (years)

41. Outcome by timing of imatinib start

42. Multivariate Cox Regression OS
Co-Variate HR
95%CI P
Imatinib
0.56
<0.0001
Age
1.02
Pres.WCC
1.32

43. OS for patients not receiving alloHSCT
100
ALL CRs but NO BMT 75 50
24% Imatinib 25
18%Pre-imatinib 1 2 3
Univariate: unadjusted p=0.08
Cox, allowing for age and WBC p=0.02
% still alive
100
Excl. Rel/died within median time to BMT 75
first including all those who achieved remission but did not receive a transplant, and secondly (in line with the ph positive paper we did) also excluding those who went off protocol in induction and those who relapsed or died before 12 weeks. On univariate analysis neither were statistically significant. 50
24% Imatinib 25
22% Pre-imatinib 1 2 3
Time in years
Univariate unadjusted: p>0.1
Cox, allowing for age and WBC p=0.05

44. TKI post alloHSCT Yes/No?
PETHEMA: post myeloablative alloHSCT poorly tolerated
only 62% started, many discontinued
U of Minn trend towards better outcome on those given imatinib
(v. small study)
Seattle RIC allo – trend towards better outcome when imatinib was
given but no effect on relapse
For everyone or selected patients ?
GMALL up-front at 3 months vs. only upon BCR-ABL re-appearance
Poorly tolerated when started early
No difference so far between the groups
For how long?
No idea….! Most studies select 1-2 years - empirically

45. Conclusions Higher CR rate by 10% with no increase in TRM
when imatinib is added to therapy
Imatinib almost doubled the rate of alloHSCT in
UKALL12/ECOG2993
Overall, significantly improved outcomes in all
endpoints measured are typical in imatinib containing
regimens
Overall Outcomes for Ph+ ALL Rx are now encouraging:
UKALL 62%
GMALL imatinib/allo
JALSG 65%

46. Conclusions
improved outcome in TKI era relates mostly to a higher rate of alloHSCT --- modest (?no) long term benefit to imatinib where HSCT not achieved
UKALL12/E2993 and GMALL studies – still poor outcome where no myeloablative alloHSCT
No role for omitting alloHSCT in adults with Ph+ ALL
Not yet clear how best to use TKI post allo

47. Strong argument for future, large international co-operative studies
I think I’d rather manage a large
international collaboration
of transplant physicians

48. Funded:
UKALL12/ECOG2993 Ph positive arm: imatinib Adele K. Fielding Georgina Buck Letizia Foroni Hillard Lazarus Mark Litzow Selina M. Luger David I. Marks Andrew K. McMillan Anthony Moorman Elisabeth Paietta Susan M. Richards Jacob M. Rowe Marty S. Tallman Anthony H.Goldstone
UK National Cancer Research Institute
Adult ALL Subgroup
USA
Eastern Co-operative
Oncology Group