1. Department of Hematopathology
Cytogenetic landscape and prognostic impact in blast phase of chronic myeloid leukemia
Shimin Hu, MD, PhD
Department of Hematopathology
The University of Texas MD Anderson Cancer Center

2. Background Chronic myeloid leukemia (CML): myeloproliferative neoplasm
BCR-ABL1/t(9;22)(q34;q11.2), variants, cryptic translocations (Ph)
triphasic natural history:
chronic phase (CP)
accelerate phase (AP)
blast phase (BP): 2/3 myeloid BP (MyBP), 1/3 lymphoid BP (LyBP)
t(9;22):
1st chromosomal abnormality identified in cancer
1st translocation molecularly characterized
1st chromosomal abnormality resulting in successful target therapy

3. Background: patient outcome
Pre-TKI (tyrosine kinase inhibitor) era:
(no treatment, radiotherapy, busulfan or hydroxyurea)
CP: median survival 3-5 years
AP: median survival 1-2 years
BP: median survival 3-4 months
TKI era:
CP: 8-year overall survival 87%;
similar to the general population if CCyR achieved
AP (initial): 8-year overall survival 75%
BP: median survival < 1 year (7-10 months)

4. Background: disease progression
Blastic transformation of CML-CP
Pre-TKI era:
most within 2-4 years;
annual incidence of >20%
TKI era:
~5%;
annual incidence of 1-1.5%
Clinical and laboratory parameters:
WBC, PLT, basophil; blast; splenomegaly
Cytogenetic and molecular genetics:
additional chromosomal abnormalities (ACAs); mutations
Treatment response parameters:
TKI resistance (ABL1 dependent or independent)
Risk factors for disease progression

5. Background: ACAs in pre-TKI era
By disease stage: CP: <10%; AP: 20%; BP: 60-80%
By ACA type:
+8 34%
+Ph 30%
i(17q) 20%
+19 13%
Major-route ACAs: >10% of all cases with ACAs (changed to 5% later)
Minor-route ACAs: others
-Y 8.8% (male)
+21 7%
+17 5%
-7 5%
Impact: major-route but NOT minor-route ACAs confer worse
prognostic impact in CP/AP; Controversial impact in BP
Johansson B, et al. Acta Haematol 2002;107:76-94

6. Background: ACAs in TKI era
Similar to pre-TKI era in frequency and pattern.?
Impact: major-route but NOT minor-route ACAs
confer worse prognostic impact in CP/AP;
Unknown impact in BP.
Management: major-route ACAs emerged during therapy indicate treatment failure; Major-route
ACAs at diagnosis don not mandate differential
treatment (ELN 2013).
Baccarani M, et al. Blood 2013; 122:

7. Unanswered questions about CML-BP in TKI era
ACA landscape in CML-BP? Similar to pre-TKI era?
Prognostic impact of ACAs in CML-BP?
Therapeutic goal in CML-BP? MMR, MR4, MUL?
Or CCyR is adequate similar to that in CMP-CP?

8. Confounding: not all ACAs created equal
All previous studies failed to take into consideration
the complexity or the risk of each individual ACAs!!!
1. 47,XY,+8,t(9;22)(q34;q11.2)[20]
2. 46,XY,inv(3)(q21q26.2),t(9;22)(q34;q11.2)[20]
Major route
Minor route
1. 46,XY,t(9;22)(q34;q11.2)[20]
2. 47,XY,+8,t(9;22)(q34;q11.2)[20]
3. 48,XY,+8,t(9;22)(q34;q11.2),+19[20]
4. 49,XY,+8,t(9;22)(q34;q11.2),+19, +der(22)t(9;22)[20]
Complex karyotype

9. Confounding factors in impact of ACAs
+8 only (n=37)
+8 with 1 additional ACAs (n=46)
+8 with 2 additional ACAs (n=35)
+8 with >2 additional ACAs (n=44) 50
100
150
200 25 75
Months O S a f t e r A C s m g n c
p<0.0001 50
100
150
200 25 75
Months O S a f t e r C M L d i g n o s
p<0.0001
Wang W, et al. Blood Jun 2;127(22):
Wang W, et al. Blood Oct 1;126(14):
Wang W, et al. Leukemia Nov;29(11):2263-6

10. Unanswered questions about CML-BP in TKI era
ACA landscape in CML-BP? Similar to pre-TKI era?
Prognostic impact of ACAs in CML-BP?
Therapeutic goal in CML-BP? MMR, MR4, MUL?
Or CCyR is adequate similar to that in CMP-CP?

11. Patient cohort Inclusion criteria
354 patients with CML-BP diagnosed from 1999 to 2015.
All patients received TKIs if diagnosed in BP; if diagnosed
in CP or AP, all patients received TKIs before blastic transformation.
Exclusion criteria
Patients with cryptic BCR-ABL1 rearrangement.
Patients with isolated myeloid sarcoma (without concurrent BP in bone marrow or peripheral blood.
Chen Z, et al. Leukemia Mar;31(3):
Chen Z, et al. Leukemia Jul;30(7):
Chen Z, et al. Blood Cancer J. 2017 Feb 3;7(2):e521
Chen Z, et al. Blood Cancer J May 6;6:e418

12. Patient cohort Demographics: 354 patients with CML-BP:
333 in CP/AP initially and BP later; 21 (5.9%) in BP initially;
214 (60.5%) men, 140 women; median age: 52 years (12-92)
Lineage of BP:
MyBP: 240 (67.8%)
LyBP: 106 (29.9%)
MPBP: 8 (2.3%)
Treatment in BP:
323 (91.2%) received TKIs (+/- chemo)
31 (8.8%) received chemo only
91 (25.7%) received SCT
Treatment response:
< HR: 217 (61.3%)
> HR: 137 (38.7%)
> CCyR: 79
> MMR: 59
MUL: 48

13. Frequency of ACAs at diagnosis of CML-BP
270/354
Chen Z, et al. Leukemia Mar;31(3):
Chen Z, et al. Leukemia Jul;30(7):

14. Distribution of ACA types in CML-BP
>10%: +8, +Ph, 3q26.2, -7, i(17q)
>5%: +21, -Y, +19

15. Complexity of ACAs in CML-BP
Simple: 68 (25.2%)
Complex: 202 (74.8%)

16. Lineage specificity associated with ACAs
LyBP: -7
MyBP: +8, 3q26.2, i(17q), +19
Both: +Ph, +21, -Y
p=.53
p<.001
p=1.0
p=.047
p=.01
p=.27

17. ACA landscape in TKI vs pre-TKI eras

18. ACA landscape in TKI vs pre-TKI eras

19. ABL1 mutation and 3q26.2 rearrangement
% for other ACAs

20. Impact of ACAs by lineage of blast phase

21. Impact of ACAs by frequency of ACAs

22. Impact of ACAs by emerging time of ACAs

23. Impact of ACAs by complexity and risk of ACAs
Complex karyotype: Ph + >2 ACAs
High-risk ACAs: 3q262.2, -7/7q-, i(17q)

24. Impact of allo-HSCT in CML-BP with ACAs

25. Relationship between ACAs in CML-BP
Pearson correlation analysis A B 1
0.283
0.025
0.225
-0.205
-0.380
0.155
0.244
0.117
-0.086
-0.114
-0.040
0.112
-0.095
-0.153
-0.169
-0.284
-0.245
-0.063
-0.096
-0.229
-0.173
Subgroup 1: 3q26, -7/7q-, i(17q)
Subgroup 2: +8, +9, +21, +Ph

26. Risk stratification of ACAs in CML-CP50
100
150
200 25 75
Months O S a f t e r A C s m g n c
3q26.2
i(17q)
+Ph -Y
-7/7q- +8
Other single
Subgroup 2
Subgroup 1
Wang W, et al. Blood 2016 Jun 2;127(22):
Wang W, et al. Blood 2015 Oct 1;126(14):

27. Treatment response for optimal outcome in CML-BP1 2 3
MUL/CMR: similar to AML with favorable cytogenetics
MMR: similar to AML with intermediate cytogenetics
CCyR: similar to AML with unfavorable cytogenetics
Chen Z, et al. Blood Cancer J. 2017 Feb 3;7(2):e521

28. SUMMARY
+8, +Ph, 3q26.2 rearrangement, -7 and i(17q) have a frequency >10% (major-route); +21 and +19 have a frequency >5%.
+8, 3q26.2 rearrangement, i(17q) and +19 are common in MyBP; -7 more common in LyBP; +Ph and +21 are equally distributed between MyBP and LyBP.
There is significant shift in the distribution of ACAs in TKI vs pre-TKI eras. The emergence of 3q26.2 rearrangement as a major-route ACA in TKI era correlates with the high frequency of ABL1 mutations.

29. SUMMARY
ACAs confer a negative prognostic value in MyBP but not LyBP. However, the nature, the frequency or the emerging time of ACAs have no prognostic relevance. The adverse impact of ACAs can be overcome by allo-SCT.
Pearson correlation analysis reveals two subgroups of ACAs. The subgrouping of ACAs correlates with their prognostic impact in CML-CP/AP, supporting their role in blastic transformation.
Differential depth of treatment response is required for optimal outcome in CML-BP vs in CML-CP (CMR vs CCyR).

30. ACKNOWLEDGEMENT Hu Lab: Department of Hematopathology
Yan Li, MD, PhD
Shi Bai, MD, PhD
Parsa Hodjat, MD
Lianghua Fang, MD, PhD
Zhihong Hu, MD, PhD
Christina Regmi, MA, MS
Vanya Jaitly, MD
Hui Liu, MD, PhD
Andy Ai, MD, PhD
Wei Wang, MD, PhD
Wen Yang, MD, MS
Wei Liu, MD, PhD
Zimu Gong, MD
Lan Zheng, MD, PhD
Zi Chen, MD, PhD
Habibe Kurt, MD
Hu Lab:
Wei Xie, MD, PhD
Arash Ronaghy, MD, PhD
Department of Hematopathology
Bone marrow laboratory
Cytogenetics laboratory
Molecular diagnostic laboratory
Department of Leukemia
CML section