Comparison of Prognostic Factors and Outcomes of Patients with Secondary Acute Myeloid Leukemia (AML) Following Myelodysplastic Syndromes (MDS), Myeloproliferative Disorders (MPD), or Therapy-Related AML (t-AML).

Background: Response rates to induction chemotherapy and survival are poorer for patients (pts) with secondary AML compared to those with de novo AML. Within the category of secondary AML, no studies have compared the outcomes of induction chemotherapy in pts with AML arising from antecedent MDS vs. from MPD vs. t-AML.

Methods: We conducted a retrospective review of all pts with newly diagnosed, pathologically-confirmed AML at a single institution between 1997 and 2005 and identified pts who also were diagnosed with an antecedent MDS, MPD, or t-AML. Data were collected on baseline characteristics and outcome, and controlled for in stepwise multivariable analyses. All pts were treated with anthracycline-based induction regimens. The primary endpoints were:

1. complete remission (CR) rate (as defined by the IWG criteria) and survival from time of AML diagnosis, to determine whether those with AML arising from MDS, from MPD, or t-AML had different outcomes; and

2. to define predictors of outcome among pts with secondary AML. Results: Of 457 AML patients, 281 were treated with remission induction therapy, of whom 66 had AML arising from MDS, MPD, or t-AML.

Thirty-one (47%) had antecedent MDS, 20 (30%) an MPD, and 15 (23%) had t-AML. Twenty-six pts (39%) were female. The median age at the time of AML diagnosis for those with MDS, MPD, and t-AML was 67, 61, and 57 years, respectively (range 36–82, p=0.03 for all). Time from antecedent diagnosis/event was 7, 42, and 38 months, respectively (p=0.001). Cytogenetic risk categories (per CALGB 8461) were favorable in 3 pts (5%), intermediate in 30 (45%), unfavorable in 22 (33%), and unknown in 11 (17%). Neither cytogenetics (p=0.19) nor FAB/WHO AML classification (p=0.43) differed among groups. Median WBC at AML diagnosis was lower for pts with AML from MDS (3.7k/uL, vs. 9.9 k/uL for AML from MPD and 7.2 k/uL for t-AML, p=0.04), as were the percentage of peripheral blasts (6%, vs. 30% for AML from MPD and 23% for t-AML, p=0.005) and incidence of splenomegaly (6%, vs. 50% for AML from MPD and 0% for t-AML, p<0.001). CR rate was not significantly lower for patients with AML from MDS (35%) than for those with AML from MPD (55%) or t-AML (53%, p=0.33). Overall median survival was 8.0 months, with no significant difference among the 3 groups (7.7 months for AML from MDS, 11.1 months for AML from MPD, and 5.2 months for t-AML, p=0.23). Gender, interval from diagnosis of the antecedent disorder to diagnosis of AML, and marrow blast count did not correlate with CR rate or survival. In multivariable analyses, the antecedent diagnosis remained non-predictive of response rate or survival. Favorable- or intermediate-risk cytogenetics, however, compared to poor-risk, were significant predictors of higher CR rates (p= .005) and longer survival (p<.001). Lower peripheral blast percentage and age <60 years were favorable prognostic factors for CR.

Conclusions: Contrary to expectations, pts with AML from MDS, from MPD, or t-AML had similar outcomes. The most important predictors of response to induction chemotherapy were cytogenetic risk group, age, and peripheral blast percentage. Only cytogenetic risk group was predictive of survival.