Efficacy and Feasibility of High-Dose Cytarabine Plus Idarubicin and Amifostine As Induction Schedule: A Prospective Observational Study of 100 AML Elderly Fit Patients

Abstract 3621

Acute myeloid leukemia (AML) has a dismal prognosis in elderly population because of intrinsic chemoresistance and frailty of patients. High-dose Cytarabine (HiDAC) in induction therapy did not improve the CR in younger AML patients and recent guidelines discourage this approach in elderly because of high extrahematological toxicity. Amifostine showed to selectively protect normal Hemopoietic progenitors from chemotherapy and we previously successfully tested the feasibility of an induction schedule including HiDAC (3 g/m² days 1,2,3,4,5), Idarubicin 40mg/m² on day 3 preceeded by Amifostine (740 mg/m²). We designed a prospective observational study including the same induction schedule, aimed to evaluate the outcome (CR rate, OS and EFS) of a larger population of fit AML elderly patients. Fit patients, selected according the Multidimensional Geriatric Assessment, received 1–2 courses and underwent PBSC mobilization after consolidation. Patients who collected ≥3×10e6CD34+/kg received ASCT, while poor mobilizers were considered for alternative regimen including Allogeneic transplantation from an HLA-matched sibling, chemotherapy (CHT) or Gemtuzumab-Ozogamicin (GO).

We registered 156 consecutive patients, aged >59 yrs; 56 were unfit for intensive induction chemotherapy and received only palliative care; 100 (64%) fulfilled the inclusion criteria of our protocol: 91 received the scheduled induction regimen, while 9 received a Fludarabine regimen because of reduced cardiac function. These patients were not included in the response evaluation, but were considered for the outcome (according to the ITT criteria). Patients' characteristics are shown in table 1.

CR was achieved in 73.6% of patients; multivariate analysis showed secondary disease as predictive of poor response, with a 65% CR rate (RR = 2.54; 95% CI: 1–6.45; p= 0.05) vs 83% in primary disease. Induction death rate was 5% and not influenced by any prognostic factors. The median time to achieve neutrophil>500× 10⁶/L and platelet>20,000×10⁶/L, were 17 and 19 days (ranges of 9–29 and 3–47 respectively). The main extrahematological toxicity were grade 3–4 mucositis (13%) and hepatic toxicity (9%). We also observed 66 grade III-IV febrile neutropenia/infectious episodes.

Overall 65 patients received a first consolidation course and mobilization for PBSC harvest; we observed 6 TRD, a 3% of grade 3–4 hepatic and neurological toxicity and 6% of grade 3–4 cardiac toxicity; in 4 patients we observed rapid early leukemia relapse; overall 55 patients were eligible for post-consolidation therapy.

Only 24 patients achieved a succesfull PBSC mobilization and ASCT was performed in 21 (2 relapsed and died before ASCT and 1 received Allogeneic Transplant). Thirty-one patients were poor mobilizers: 3 received Allogeneic Transplant, 3 CHT, 5 stopped treatment because of persistent aplasia and 20 received low-dose GO (3 mg/m2 monthly for 3 times and every 3 months after; median: 3, range 3–6 courses).

With a median follow-up of 70 months (range 24–124) 21 patients are alive (19 in continuous CR), 6 after ASCT, 13 after GO, 1 after CHT.

The 8 yrs Overall Survival (OS), Disease Free Survival (DFS) and Event Free Survival (EFS) are respectively 20.4% (median: 11.4 months), 24,3% (median 8.8 months) and 17,7% (median: 8.8 months).

Secondary AML and hyperleukocytosis are factors predictive of OS at the multivariate analysis. Patients with secondary disease have a 1.59 RR to die with a 9.9% 8 yrs OS vs 27.1% of patients with primary AML. Patients with WBC ≥50,000/mcl had a 2.2 RR to die with a 0% OS at 33 months vs 23.2% 8 yrs OS in patients with WBC<50,000/mcl.

In conclusion our novel intensive induction regimen for fit AML patients is safe and effective both in term of CR rate and outcome. The ASCT feasibility was confirmed to be poor in this setting (21%) while GO low-dose seems to be feasible and promising. Finally our prospective study in 156 elderly AML patients describes the real-life outcome of this setting, suggesting that two thirds of AML elderly patients are fit for intensive treatment and that long term OS can be achieved in a relevant proportion of patient with de novo AML.