Abstract
The DNA hypomethylating agents 5-azacytidine (aza) and decitabine are a new treatment option for AML/MDS pts relapsing after allografting. Cancer/testis antigens are upregulated in vitro and in vivo by these drugs (Goodyear et al., Blood 2010, Almstedt et al., Leuk. Res. 2010), hypothetically sensitizing (“priming”) malignant cells to DLIs. In 2006 we initiated a program of administering aza (100 mg/day) for 3 consecutive days, followed whenever possible by DLI 7 days after the last aza injection (repeated every 3-4 weeks). Of 26 pts treated thus, more than 50% achieved at least temporary disease control, with 4 pts achieving complete remissions (CRs) of >12 months duration at time of reporting (Lübbert et al., BMT 2009). We now updated the results of these 26 pts and report on the further 46 pts treated on this program.
Between 10/2006 and 05/2013, 72 pts (median age 62 years, range 20–75, 46 male, 26 female) with hematological relapse after allografting were treated with aza and, if at all feasible, with DLI as described above. A 5-day schedule of 100 mg/m2 aza daily as starting dose was introduced in 2009 for pts with higher leukemic load. Diagnoses were AML (n=62), MDS (n=6), CMMoL (n=2), acute mast cell leukemia and blastic plasmacytoid dendritic cell neoplasm (1 each). Of 68 pts with informative cytogenetics, 44 had intermediate-2 or adverse cytogenetics according to ELN Guidelines. Sixty pts had received fludarabine-based reduced-toxicity conditioning regimens (mostly FBM), 12 had received myeloablative conditioning; 23 pts had a HLA-matched sibling donor, 47 pts an unrelated donor and 2 pts a syngeneic sibling donor. Median duration from transplant to first day of aza was 289 days (range, 59 to 2133), 48,19 and 2 pts were in 1st, 2nd or 3rd hematologic relapse, respectively, at start of aza. Two pts had only mixed chimerism, one had only extramedullary disease at 1st relapse. Median % of blasts before aza was 4 % (range 0 to 91) in peripheral blood and 25 % (range 0 to 100) in bone marrow. Response criteria (for CR, temporary disease control or treatment failure) were applied as previously described. Uni- and multivariate analyses were performed.
47 pts (65 %) and 18 pts (25 %) received a 3-day or 5-day course of aza, respectively, as starting dose. In 3 (4 %) and 4 pts (6 %), initial dosing was limited to 2 and 4 days, respectively. A total of 189 courses of aza were administered (median of 2.7 courses, range 1 to 10). 65/72 pts (90 %) also received DLIs, 41/72 (57%) already prior to first aza. Only 7 pts developed acute GvHD after DLI treatment. Clinically relevant neutropenic infections not associated with disease progression developed in 4 pts, in one of them leading to fatal sepsis. 39 pts (54 %) benefited from the treatment, with CR in 7 pts (9.7%), in 2 of them now lasting more than 5 yrs. Thirty-two pts (44 %) had temporary disease control with stable mixed chimerism (median duration 71 days, range 31-380), with 2 pts converting to complete donor chimerism lasting 146 and 380 days, respectively. The median survival from start of aza was 108 days. Nine pts are alive at the time of this report. Twenty-one pts proceeded to subsequent transplant, 4 of them (19%) are presently alive. Among 10 predictive factors, 5 were significant at p<0.1 and thus subjected to a multivariate analysis (Cox model; Table 1). Peripheral blood blasts <1 % retained their predictive value for better overall survival (32.7 vs. 14.7 weeks, p=0.03).
. | . | HR . | 95% CI for HR . | p-value . |
---|---|---|---|---|
DLI prior to first 5-azacytidine | Yes | 1 | - | 0.09 |
No | 1.61 | [0.92, 2.81] | ||
blood stem cell donor | Unrelated | 1 | - | 0.18 |
Related | 0.62 | [0.31, 1.25] | ||
weeks from transplant to relapse | >=36 | 1 | - | 0.08 |
<36 | 1.72 | [0.94, 3.15] | ||
peripheral blood blasts | <1% | 1 | - | 0.03 |
>=1% | 1.98 | [1.07, 3.66] | ||
Donor age | <=40 years | 1 | - | 0.83 |
>40 years | 0.93 | [0.50, 1.75] |
. | . | HR . | 95% CI for HR . | p-value . |
---|---|---|---|---|
DLI prior to first 5-azacytidine | Yes | 1 | - | 0.09 |
No | 1.61 | [0.92, 2.81] | ||
blood stem cell donor | Unrelated | 1 | - | 0.18 |
Related | 0.62 | [0.31, 1.25] | ||
weeks from transplant to relapse | >=36 | 1 | - | 0.08 |
<36 | 1.72 | [0.94, 3.15] | ||
peripheral blood blasts | <1% | 1 | - | 0.03 |
>=1% | 1.98 | [1.07, 3.66] | ||
Donor age | <=40 years | 1 | - | 0.83 |
>40 years | 0.93 | [0.50, 1.75] |
In an extended cohort of pts we could confirm our previous results of clinical activity of this approach. Notably, there are long-term remitters after this well-tolerated outpatient treatment, which appears particularly effective in pts without peripheral blood blasts. Recurrent responses to the aza / DLI combination in pts who had progressed under DLI-only treatment support the “priming” concept whereby hypomethylating agents can sensitize malignant cells to other biological or cellular therapies, thus acting as so-called biological response modifiers.
Off Label Use: 5-azacytidine, immunomodulation.
Author notes
Asterisk with author names denotes non-ASH members.