Geoffrey P. Herzig, MD 1941-2013

Allogeneic Hematopoietic Stem Cell Transplantation (AlloHSCT) for Relapsed and Refractory (R & R) Acute Myeloid Leukemia (AML) An sin of omission…or commission?* *With apologies to Plutarch

Full quote… “The omission of good is no less reprehensible than the commission of evil.”

Plutarch (Lucius Mesterius Plutarchus), AD 46-120

Briefly, some background • Failure to reach CR-1/relapse is the main cause of failure after any/all therapies for AML.

SEER Stat Fact Sheets: Acute Myeloid Leukemia (AML)

More background • Failure to achieve CR/relapse is the main cause of failure after any/all therapies* for AML. • AML is the chief indication for alloHSCT in the USA, and most patients are transplanted in CR-1.

*Including AlloHSCT!

Percent of AML patients who received a first alloHSCT between 2000 and 2013 (CIBMTR, 2015) 1st CR

Other CR

Relapse

PIF

Untr/Msg

60% (1616)

50% 40%

30% 20%

(529) (406)

10% 0%

(210)

Even more background • Failure to reach CR-1/relapse is the main cause of failure after any/all therapies for AML. • AML is the chief indication for alloHSCT in the USA, and most patients are transplanted in CR-1. • The alloHSCT “toolbox” includes escalated-dose cytotoxic therapy and/or GvL

Some acknowledged limitations in the literature… • Selection (and probably publication) bias is profound - Fragile clinical status - Donor acquisition issues • Anything beyond the terms “primary induction failure” (PIF) and “relapse” (e.g., CRi, CRc, MRD) is not detailed consistently • Even excluding APL, AML is a heterogeneous disease, and only the most recent reports include molecular (e.g.) data • BMT reports often mix such patients are with those with “high-risk” features who are in CR* *Not ideal, but…not as bad as it seems?

Goal: Answer the question! A sin of…

Omission?

• Patient selection • Optimal approach • Improvements (> Saturday PM)

Commission?

STOP!

A few key questions…

Q1: Is there a (differential) cure rate with alloHSCT in “R & R” AML?

*Both AML and ALL

Thomas, et al., Blood 1977

Quantifying the Survival Benefit of AlloHSCT in “R & R” AML (at MDACC) 599 AML patients identified at first salvage

67 patients died during first salvage 42 (8%) HSCT as first salvage 490 (92%) patient underwent first salvage chemotherapy

112 (23%) remission

30 (71%) Remission

12 (29%) No remission

378 (77%) No remission 94 patients received no further treatment

112 underwent postremission therapy

46 (42%) HSCT

66 (58%) Chemotherapy

284 underwent additional salvage therapy

84 (22%) HSCT

200 (53%) Chemotherapy

Armistead, et al., Biol Blood Marrow Transplant. 2009

Quantifying the Survival Benefit for Allogeneic Stem Cell Transplantation in “R & R” AML

Armistead, et al., Biol Blood Marrow Transplant 2009

Q2: Are we getting better at curing these patients?

*Both AML and ALL

Thomas, et al., Blood 1977

Q3: Can we select patients destined to do better/worse with current techniques?

Patient selection…

Hematopoietic Stem-Cell Transplantation for Acute Leukemia* in Primary Induction Failure or Relapse • Duration of CR1 < 6 months • Circulating blasts • Donor other than MSD/MUD • KPS < 90% • Poor risk cytogenetics

*Both AML and ALL, AML shown

Duval, et al., J Clin Oncol, 2010

CIMBTR Analysis of BMT in Relapsed AML and ALL

Duval, et al., J Clin Oncol, 2010

Other factors of importance… • Donor availability • Performance status, etc. • HCT-CI score • Molecular studies • Patient’s choice • Programmatic issues

Q4:

In a transplant-eligible patient, how does one decide whether or not to re-induce R & R patients?

Induction Failure (PIF)

AlloHSCT as Initial Therapy in AML patients primarily refractory to HiDAC

Jabbour, et al., Am J Hematol 2014

Relapse

Prognostic Index for Adult Patients with AML in First Relapse

Breems, et.al., J Clin Oncol 2005

“Breems Index” (CR rates by group) A = Favorable (1-6): 85%

B = Intermediate (7-9): 60% C = Unfavorable (10-14): 34% All = 46%

Evaluation of the impact of remission induction chemotherapy prior to allogeneic stem cell transplantation in relapsed and poor-response patients with AML (ETAL3-ASAP)

Relapsed AML

Poor response AML

An Alternate Approach

FLAMSA/RIC ± DLI

Schmid, et al., JCO 2005

FLAMSA-RIC for Allo SCT in “R & R” AML Median Follow up (survivors):

Overall survival Leukemia free survival

36.5 (16-76) mo

Overall Survival

n = 103

1 year 50% 2 years 39% 4 years 37%

Schmid et al., Blood 2006, updated

FLAMSA-RIC: Superior Results in PIF vs. Relapsed Disease Allo SCT in refractory AML in general...

Schmid et al., Blood 2006

... And in CN-AML

Pfeiffer et al., et al., Haematologica 2012

Summary of FLAMSA/RIC  AlloHSCT ± DLI • Patients: >800 in 13 publications, ‘05-’16, all with various poor prognostic signs (but not all “R&R”) • Results: CR > 70+%, DIA ~ 10%, NRM2 ~ 20-30%, RI2 ~ 40-50%, DFS2 ~ 35%, OS2 ~ 40-45% • Prognostic features: The usual suspects, but notably, those with PIF, “fewer” blasts and less prior therapy have improve outcomes • Works in progress: Amsacrine replacement(s), RIC substitutions (i.e., Bu, Mel, TT vs TBI), early DLI

Can we improve intrinsic results? • Conditioning • Donor selection/graft engineering • GvL manipulation (DLI, etc.) • Maintenance

An Ideal Maintenance Agent • Active against AML • Not too toxic/can be given early after transplant • Increase immunogenicity of malignant cells • Influences GvL

Protocol MDACC 20080503/NCT00887068 • Hypothesis: Low-dose 5-Azacitidine maintenance will decrease the relapse rate after alloHSCT for “high risk” AML/MDS • Study aim: Randomized comparison of 1 year maintenance with low-dose AZA 32 mg/m2 daily x 5 days, in 28-day cycles, for 1 year, vs. nil • Start/stop dates: April 09, 2017; as of January 2015, n=153 patients

• Goal: Detect a 50% prolongation in EFS

Summary* • Clinical trial preferred (necessary?) • Choose wisely…and quickly. (Planning helps!) • Select donor – based on prompt availability • Use MAC > RIC > NMA • Consider the “FLAMSA, etc.” approach • Test post-transplant maintenance therapy (-ies?) *If the answer is (often!) “a sin of omission”

RESERVE SLIDES

How low can you go?

FHCRC NMA and AML < CR*

Gyurkocza, et al., JCO 2010

ETAL-3 ASAP “Evaluation of the impact of remission induction chemotherapy prior to allogeneic stem cell transplant in relapsed or poor-response patients with AML” • Multicenter, randomized • Endpoint is D +56 survival • Inclusion criteria - Age 18-75 - Suitable… • Exclusion - APL - Usual…

Donor Considerations • Timing is critically important: MSD, (D-) UCB and PMRD (haploidentical) > MUD or PMUD • If DLI is a consideration: MSD/MMRD > MUD >>> D-UCB

• No clear advantage regarding GvL/GvHD with these donors? • Thus, “winners” = MRD, PMRD?

Cause of Death by Disease Status After HLA-Identical Sibling Transplants for AML done in 2011-2012

Cause of death

Early disease (%)

Intermediate disease (%)

Advanced disease (%)

Unknown Total (%)

Primary disease

277 (57)

58 (41)

188 (65)

7 (70)

530

GVHD

68 (14)

30 (21)

35 (12)

1 (10)

134

Infection

61 (13)

15 (11)

31 (11)

1 (10)

108

Organ failure

14 (3)

10 (7)

9 (3)

0

33

Secondary malignancy

4 (1)

1 (1)

0

0

5

Other

60 (12)

28 (20)

27 (9)

1 (10)

116

Total

484

142

290

10

926

CIBMTR, 2015

Some patients should just be avoided…

Donor Leukocyte Infusions (DLI): Cliff Notes • First reported in CML (Kolb, et al., Blood 1990) • Less potent in AML (Takami, et al., BBMT 2015)

• Variably useful in reversing, but more helpful in preventing relapse? (Wang, et al., BMT 2012) • Plagued by frequent, often severe/fatal GvHD – especially if given < D+100 or in patients with prior GvHD (de Lima, et al., BMT 2001) • Better in combination with “chemotherapy”? (Schroeder, et al., BBMT 2015)

Manipulating GvL/GvHD* • Donor selection • Avoidance/truncation of GvHD prophylaxis

• Donor leukocyte infusions (DLI) • Others…

*It’s been tried

Classification of conditioning regimens in 3 categories

A. Bacigalupo, et al., BBMT 2012

Allogeneic marrow transplantation in patients with AML in CR1: A randomized trial of two TBI dose regimens

Clift, et.al., Blood 1990

Reduced-intensity and myeloablative conditioning allogeneic hematopoietic stem cell transplantation in patients with acute myeloid leukemia and myelodysplastic syndrome: a meta-analysis and systematic review

Forest plots from a meta-analysis of transplantation outcomes for RIC compared to MAC arm. Wen Zeng, et al., Int J Clin Exp Med 2014;7(11):4357-4368

Some things I won’t (but perhaps should?) discuss regarding conditioning… • Older (≥ 60-65 years), fit patients do well with certain “fullintensity” regimens (Alatrash, et al., BBMT 2011)

• Escalated versions of “legacy” regimens exist, but have not been fully accepted (Russell et al., BMT 2008) • Addition and/or substitution of chemotherapy drugs (e.g, clofarabine, carmustine or treosulfan) to current regimens (Chevallier, et al., Haematologica 2014; Slack, et al., BBMT 2013; Gyorkocza, BBMT 2014)

• Addition of radioimmunotherapy (Mawad, BBMT 2014)

Can we improve intrinsic results? • Conditioning • Donor selection/graft engineering

Can we improve intrinsic results? • Conditioning • Donor selection/graft engineering • GvL manipulation

Maintenance therapy with low-dose azacytidine after allogeneic stem cell transplantation for recurrent AML/MDS: A dose and schedule finding study • 45 high-risk AML patients, median age 60 years, 67% < CR.

• 5 daily doses (8-40 mg/m2), D +40, then monthly cycles, 1-4 • Optimal dose: 32 mg/m2 x 4 cycles • One-year EFS/OS % = 58/77 • Basis for RCT de Lima, et.al., Cancer 2010

Mortality by Disease Status After HLA-Identical Sibling Transplants for AML done in 2011-2012 Unknown, 1.1%

Advanced, 31.3% Early, 52.3% Intermediate, 15.3%

The data presented here are preliminary and were obtained from the Coordinating Center of the Center for International Blood and Marrow Transplant Research. The analysis has not been reviewed or approved by the Statistical or Scientific Committees of the CIBMTR.

An integrated approach

Variants of the Sequential Therapy Concept Pilot trial Munich / Augsburg: FLAMSA – BU N = 42, median age : 65.3 y FLAMSA-Busulfan n = 29,1y OS 69,6% FLAMSA-TBI, 1 y OS 41,6%

- FLAMSA-BU-FLU - FLAMSA-Treo-FLU

- FLAMSA-MAC… FLAMSA-(TT-)Mel - Clofarabin-AraC (CLARA) instead of FLAMSA

BMT CTN 0901 PROTOCOL: A randomized, multi center, phase III study of Allogeneic Stem Cell Transplantation comparing regimen intensity in patients with Myelodysplastic Syndrome or Acute Myeloid Leukemia Reduced Intensity Conditioning (RIC)

Myeloablative Conditioning (MAC)

A Fludarabine/Busulfan (Flu/Bu)

C Busulfan1/Fludarabine (Bu/Flu)

B Fludarabine/Melphalan (Flu/Mel)

D Busulfan1/Cyclophosphamide (Bu/Cy)

• Fludarabine (120-180 mg/m2) • Busulfan (≤ 8 mg/kg PO or 6.4 mg/kg IV)

• Fludarabin (120-180 mg/m2) • Melphalan (≤ 150 mg/m2)

• Busulfan (16 mg/kg PO or 12.8 mg/kg IV) • Fludarabine (120-180 mg/m2)

• Busulfan (16 mg/kg PO or 12.8 mg/kg IV) • Cyclophosphamide (120 mg/kg)

E Cyclophosphamide/Total Body Irradiation (Cy/TBI) • Cyclophosphamide (120 mg/kg) • TBI (1200-1420 cGy)

Breems Index

Breems Index III

AlloHSCT with reduced-intensity conditioning following FLAMSA for primary refractory or relapsed AML Schneidawind, et al., Ann Hematol. 2013 Oct;92(10):1389-95.

Can we optimize GvL without increasing GvHD?

Tyrosine kinase inhibitors in Ph positive disease 1. Active against the disease 2. Not too toxic 3. Not myelotoxic (or with tolerable myelotoxicity) 4. Can be given early after transplant

Prognostic Index for Adult AML in First Relapse

Breems, et.al., J Clin Oncol 1995

Breems Index II

Definitions • CR: Bone marrow blasts 1.0 x 109/L (1000/µL); platelet count >100 x 109/L (100,000/µL); independence of red cell transfusions • CRi: All CR criteria except for residual neutropenia ( 3 plus intermediate and high disease risks. NRM, nonrelapse mortality; OS, overall survival; RFS, relapse-free survival. Sorror, et.al., J. Clin Oncol 2007

Why this is important… • Lingering questions about the role of alloHSCT in AML-CR1 • Continued, currently static (?) failure rate of conventional therapy in AML • Difficulty in getting relapsed patients into CR2 and ready/fit for transplant • Prompt availability of alternative donors

• Availability of newer agents

The Yin and Yang of alloHSCT for AML

Low dose azacitidine to treat relapsed AML/MDS after allogeneic transplant • Relapsed AML/MDS after allogeneic HSCT

• Doses of 16-40 mg/m2 for 5 days in 28-30-day cycles induced complete remission and reversion to full donor chimerism in 20-25% of patients treated (n=19) Jabbour et al., Cancer 2009;115(9):1899-1905

An ideal maintenance agent: Vidaza? • Active against the disease (Ramos, et al., Leukemia 15) • Not too toxic • Can be given early after transplant • Increase immunogenicity of malignant cells • Influences donor cells favorably

Hypomethylating agents – potential GvL effects • Increased expression of tumor-associated antigens I

(Roman-Gomez, 2007; Tatjana Stankovic et al.; Goodyear et al.)

• Increased expression of KIR ligands on hematopoietic cells (Liu, 2009)

• Recovery of reduced expression of HLA class I, II, and III antigens on tumor cells (Campoli & Ferrone, 2008; Pinto, et al., 1984)

• Increased expression of known minor antigens (Hambach, 2009)

• Affect microRNA function – inhibition of oncogenes

Hypomethylating agents: amelioration of GvHD? • A pharmacologic apoptotic effect on alloreactive T cells (Sánchez-Abarca et al., 2010)

• Alloreactive donor T cells are converted to immunosuppressive regulatory T cells (CD4+CD25+Foxp3+) through Foxp3 promoter hypomethylation (Choi, et al., 2010) • T regs already present in the graft contribute to the antiGvHD effect (Cooper – DePersio; ASH, 2013)

FLAMSA/RIC ± DLI

FLAMSA/RIC Issues Negatives: • Prolonged hospitalization/increased costs (?) • M-amsacrine not available in US • Donor must (?) be available when starting Positives: • Obviates the decision regarding salvage chemotherapy vs transplant • May apply to other groups of high-risk patients

Conditioning regimen schematic

Warren & Deeg, Tissue Antigens, 2013

Effect of chronic GvHD on key outcomes Relapse

NRM

DFS

Condition according to the new severity score in the IBMTR-1 cohort. Blue (a) indicates low risk; green (b) intermediate risk; red (c) high risk; and black (d), no chronic GVHD. Lee, Blood 2002

Causes of Death After HLA-Identical Sibling Transplants for AML, done in 2011-2012 Organ failure, 3.6%

Other, 12.5%

Infection, 11.7%

GvHD, 14.5%

CIBMTR, 2015

Secondary malignancy, 0.5%

Primary disease, 57.2%