Preliminary Results of a Phase II Study of Low Dose Decitabine as a Single Agent in Older Patients (age≥60) with Previously Untreated Acute Myeloid Leukemia (AML)

The hypomethylating agent decitabine inhibits DNA methyltransferase (DNMT) enzymatic activity and is approved by the FDA for treatment of myelodysplastic syndromes. We previously performed a phase I trial in AML of decitabine (with or without the histone deacetylase inhibitor valproic acid); the trial was designed to establish a biologically effective dose (BED) of decitabine based on drug-induced re-expression of methylated and silenced genes. With the derived BED of 20mg/m²/day given for 10 days, every 4 weeks, clinical activity was observed in previously untreated older AML patients (pts, CR in 4/12); clinical response correlated with gene re-expression (Blum, JCO 2007). The addition of tolerable doses of valproic acid did not appear to improve clinical results or increase gene re-expression compared to decitabine alone. Therefore, we designed a phase II study of single agent decitabine for previously untreated AML pts of age≥60 who were not candidates for intensive chemotherapy (or who refused it). Performance status was ECOG <3. All pts received induction with decitabine at 20mg/m² IV over 1 hour on days 1–10 of a 28 day cycle. The schedule for subsequent cycles was customized for each patient based on clinical response and/or toxicity (e.g., myelosuppression). Pts with persistent AML at the end of a cycle received a repeat of the 10 day treatment. In contrast, pts with <5% blasts received decitabine as consolidation for only 5 days/cycle (every 4 weeks for 1 year). For CR/CRi patients, the decitabine schedule could be reduced further to 4 or 3 days/cycle for Grade 4 neutropenia lasting more than 2 weeks, if applicable. 33 pts were enrolled; accrual was completed in 13 months. Pts had a median age of 74 years (range, 60–83); 24 pts were age≥70. 18 pts had de novo AML; 15 had secondary or t-AML. Pts had a median presenting white blood cell count of 2,400/uL (range, 400–58,800/uL) with median marrow blasts of 44% (range, 21–92%). 13 pts had complex karyotype (≥3 abnormalities); 12 had normal karyotype; one patient had t(8;21), and the rest had other abnormalities (with one unknown). This was a high risk group by any comorbidity measure. Scoring was as follows on four high-risk factors of age ≥70, antecedent hematologic disorder, unfavorable karyotype, and ECOG 2: 16 pts had 3+ risk factors; 15 had two, and the other 2 pts had one risk factor. Using the transplant comorbidity index for older AML pts (HCT-CI; Giles, BJH 2007), pts had scores of 3+ (N=18), 1–2 (N=10), or 0 (N=5). The median number of cycles received to date is 3 (range, 1–10); 17 pts continue to receive study therapy. 10 pts have received 4+ cycles. Of the first 22 consecutive pts enrolled, 11 achieved CR (50%); 4 more have achieved CRi by IWG criteria for response (Cheson, JCO 2003). 11 pts are <3 months since entry onto the study, with response evaluations ongoing. The response rate was similar between cytogenetic risk subgroups and in both de novo and secondary AML. 4 pts who achieved CR subsequently relapsed, with CR duration of 9, 6, 6, or 2 months, respectively. The other CR pts have maintained ongoing responses for 2–11+ months. 9/11 pts who achieved CR required only 1 cycle of induction therapy (10 day course) before initial response was achieved (CR, N=2; CRi, N=7). For pts who had CRi as initial response, the median number of additional cycles to achievement of full CR was 1 (range, 1–3). In this high risk group, death within 8 weeks of any cause occurred in 4 pts and was related to infection in each case. Though therapy was often administered in the outpatient clinic, febrile neutropenia and infections in the setting of drug and/or disease-related myelosuppression requiring hospitalization during the induction courses were frequent, occurring in nearly all pts. Post-CR therapy, however, was very well tolerated with no hospitalizations for complications during consolidation courses. In conclusion, decitabine in this novel schedule of induction with outcome-adapted modification of consolidation therapy was highly active and well tolerated by most in this poor risk cohort of older AML pts. Clinical results compare favorably to those seen in this population with low dose clofarabine (Erba, ASCO 2008a) or decitabine-5 day schedule (Cashen, ASH 2006a). Correlative studies including re-expression of epigenetically silenced genes, target gene promoter methylation, and global DNA methylation are ongoing.