Abstract

Abstract 2804

Background:

The “monosomal karyotype” cytogenetic subgroup (MK+), as defined by Breems et al. (1) encompasses mostly complex karyotypes (CK, >3 abnormalities) and is associated with very poor outcome in AML (1, 2) and MDS (3). Discussions are ongoing regarding treatment response prediction, and a prognosis-modifying effect within the CK+ AML subgroup (1, 4). In a study of 227 AML pts >60 years, 38 of whom were MK+, we found that particularly pts with at least 2 monosomal autosomes (MK2+, by definition also CK+) appeared to benefit from DAC (5). We now applied this question to higher-risk MDS pts randomized to either DAC or Best supportive care (BSC) in the 06011 trial (6) which explicitly enrolled pts with IPSS poor-risk cytogenetics.

Methods:

Of 233 randomized pts, 206 had cytogenetics informative for MK status: 63 had normal cytogenetics (CN), 143 had cytogenetic abnormalities (CA) without MK (MK-, n=73) or with MK (MK+, n=70). This last group was further subdivided into pts with 1, 2 or at least 3 monosomal autosomes (MK1, MK2, MK3+, n=17, 22, 31, resp.). Endpoints were overall survival (OS), progression-free survival (PFS) and response rate (RR, complete and partial remissions, hematologic improvement). Analysis was based on the intent-to-treat principle.

Results:

As recently published (6), in the overall 233 pts, PFS was significantly improved in the DAC arm as compared to BSC (median 0.55 vs 0.25 years, hazard ratio [HR] 0.68, p=0.004), but not OS (HR 0.88), probably due to post-progression treatments, suboptimal DAC schedule and treatment duration as possible confounders. In the 206 pts with informative cytogenetics, significant improvement in outcome with DAC vs. BSC was also seen for PFS (p=0.022, HR 0.72, Table 1) but not OS (HR 0.93). The improvement of PFS with DAC vs BSC was quite pronounced in the 63 pts with CN (HR 0.55, p=0.03), but less impressive and not significant in the 143 pts with CA (HR 0.76, p=0.11). When subdividing the latter by MK categories, pts with either 2 (MK2) or more than 2 (MK3+) monosomal autosomes also had significantly prolonged PFS with DAC, reflected in RR of 67% in MK2 and 33% in MK3+ DAC-treated pts (Table 1). This effect was not obvious in the MK1 subgroup, where the PFS outcome was favorable in both treatment groups, and in the MK− subgroups, either without (“MK−/CK−”) or with complex karyotype (“MK−/CK+”), the latter having a very poor outcome in both treatment groups. With caution (considering the limited size of some subgroups), these results support our observation that AML pts with a complex karyotype harboring 2+ monosomies can gain benefit from this hypomethylating agent.

Conclusions:

This first randomized trial addressing the predictive value of the MK genotype in the presence of 2+ monosomies embedded in a complex karyotype demonstrates a very rapid deterioration of MDS pts receiving BSC (within <6 weeks from randomization with 3+ monosomies). DAC treatment mitigated the negative effect of these karyotypes. We speculate that this effect is linked to the distinct mechanism of action of the hypomethylating agent DAC (7), and that with a different DAC schedule and longer treatment duration, these effects might be observed also on OS.

DACBSC
Groupnnmedian PFS (years)% RRnmedian PFS (years)% RRp value*HR** [CI]***
all 206 105 0.50 33 101 0.25 0.022 0.72 [0.54; 0.94] 
all (#)        0.003 0.63 [0.46; 0.85] 
CN 63 36 0.86 36 27 0.38 0.034 0.55 [0.26; 1.15] 
CA 143 69 0.45 32 74 0.25 0.11 0.76 [0.49; 1.18] 
MK- 73 30 0.36 17 43 0.26 0.41 0.81 [0.43; 1.55] 
MK-/CK- 54 21 0.52 19 33 0.26 0.22 0.70 [0.33; 1.48] 
MK-/CK+ 18 0.30 11 0.23 0.92 0.95 [0.26; 3.49] 
MK+ 70 39 0.47 44 31 0.21 0.20 0.73 [0.38; 1.38] 
MK1 17 0.71 44 0.74 0.81 0.88 [0.24; 3.24] 
MK2 22 0.67 67 13 0.23 0.047 0.41 [0.13; 1.34] 
MK3+ 31 21 0.37 33 10 0.11 0.046 0.44 [0.15; 1.30] 
MK2+ 53 30 0.45 43 23 0.18 0.016 0.50 [0.23; 1.06] 
MK2+ (#)        0.006 0.43 [0.19; 0.95] 
DACBSC
Groupnnmedian PFS (years)% RRnmedian PFS (years)% RRp value*HR** [CI]***
all 206 105 0.50 33 101 0.25 0.022 0.72 [0.54; 0.94] 
all (#)        0.003 0.63 [0.46; 0.85] 
CN 63 36 0.86 36 27 0.38 0.034 0.55 [0.26; 1.15] 
CA 143 69 0.45 32 74 0.25 0.11 0.76 [0.49; 1.18] 
MK- 73 30 0.36 17 43 0.26 0.41 0.81 [0.43; 1.55] 
MK-/CK- 54 21 0.52 19 33 0.26 0.22 0.70 [0.33; 1.48] 
MK-/CK+ 18 0.30 11 0.23 0.92 0.95 [0.26; 3.49] 
MK+ 70 39 0.47 44 31 0.21 0.20 0.73 [0.38; 1.38] 
MK1 17 0.71 44 0.74 0.81 0.88 [0.24; 3.24] 
MK2 22 0.67 67 13 0.23 0.047 0.41 [0.13; 1.34] 
MK3+ 31 21 0.37 33 10 0.11 0.046 0.44 [0.15; 1.30] 
MK2+ 53 30 0.45 43 23 0.18 0.016 0.50 [0.23; 1.06] 
MK2+ (#)        0.006 0.43 [0.19; 0.95] 
*

PFS comparison: result of the logrank test, unstratified or stratified (#) by MK status;

**

HR DAC vs BSC, estimated via the Cox model, unstratified or stratified (#) by MK status;

***

CI, Confidence interval of the HR, either 95% (overall comparison) or 99% (all subgroup analyses).

Disclosures:

No relevant conflicts of interest to declare.

References:

References:
1.
2.
3.
4.
5.
6.
7.

Author notes

*

Asterisk with author names denotes non-ASH members.