Between September 2001 and October 2011, 91 newly diagnosed patients with de novo AML in the Institute of Hematology and Blood Diseases Hospital (Tianjin, China) were enrolled in the present study. Patients with acute promyelocytic leukemia were excluded. Absence of heart, lung, liver, or kidney damage, as well as no active infection, was required for enrollment in the study. Written informed consent was obtained from each patient. The diagnosis was based on bone marrow cell morphology, cytochemistry, immunophenotyping and chromosomal analysis. The final follow-up was March 2014. For the entire cohort, the median follow-up time was 46 months.

The diagnosis and classification of AML were based on the criteria of the French-American-British (FAB) Cooperative Study Group (11) and the more recent World Health Organization classification (12). The disease status was assessed by bone-marrow cytology and flow cytometric analysis (FACSCalibur; Becton Dickinson, San Jose, CA, USA). Bone marrow or peripheral samples were harvested following 1–3 days of culture, containing no stimulating factors. Chromosomal analysis was performed using the R-banding method and karyotyped, according to the International System for Human Cytogenetic Nomenclature (13). Mutations in several genes, including NPM1, FLT3/ITD, FLT3/TKD, AML1, CEBPA, C-KIT and JAK2, were investigated. For cytogenetic analysis, the chromosome karyotype and/or fusion gene (e.g. AML1/ETO or CBFβ/MYH11) mutation was analyzed successfully in 85/87 patients. A total of 85 patients were divided into three cytogenetic risk groups, according to the Southwest Oncology Group (SWOG) criteria: Favorable group with 28 cases, including t(8;21) and inv(16); intermediate group with 45 cases, including normal karyotype, 2Y, 18, 16, del(12p); adverse group with 12 cases, including 25/27, abnormal 3q, 9q, 11q, 21q, 17p, t(6;9) and complex karyotype. Since the prognosis of the unknown group was similar to that of the intermediate group, these patients were incorporated into the intermediate group.

The initial induction regimen used in patients with primary induction failure (n=44) included cytarabine (100 mg/m² infusion) for 7 consecutive days and daunorubicin (45 mg/m²) for 3 consecutive days, cytarabine (100 mg/m² infusion) for 7 consecutive days and Idamycin (12 mg/m²) for 3 consecutive days, and homoharringtonine (2–2.5 mg/m² infusion) for 7 consecutive days, a standard dose (sd) of cytarabine (100 mg/m²) for 7 consecutive days and daunorubicin (45 mg/m²) for 3 consecutive days. Daunorubicin or idarubicin, combined with 2–2.5 mg m⁻² ·d⁻¹ homoharringtonine on days 1–7 and a sd or intermediate dose of cytarabine (1 g m⁻² q12 h on days 5–7) was also performed as a first induction regimen.

The MAC regimen consisted of 8 mg m⁻² ·d⁻¹ mitoxantrone on days 1–3, 100 mg m⁻² ·d⁻¹ cytarabine on days 1–7 and 400 mg m⁻² ·d⁻¹ cyclophosphamide on days 2 and 5.

The efficacy was established, according to the criteria of Cheson et al (14): i) CR, ≤5% blasts infiltration in bone marrow without Auer's rodes, with ANC >1×10⁹/l and PLT >100×10⁹/l; ii) Partial remission, 5–25% of blast infiltration and blast reduction by 50% in bone marrow or <5% of blasts in the bone marrow, however, with Auer's rods, and with a CR criteria for peripheral blood; iii) Non-remission, the patients who succumbed to mortality prior to the assessment of remission or those who failed to achieve CR or PR; iv) Early mortality was defined as mortality from any cause during the first 30 days from the beginning of the induction treatment. Patients were considered relapsed when they had >5% blasts in the bone marrow examination.

Disease-free survival (DFS) was calculated from the date of CR to the date of relapse, mortality, undergoing stem cells transplantation (SCT) or the final follow-up. The OS was calculated from the first day of MAC administration until mortality, undergoing SCT or the final follow-up.

Survival distributions were estimated by the Kaplan-Meier method. Statistical differences in survival distributions were evaluated using the log-rank test. Data analysis was performed using R 2.10.1 software (R Foundation for Statistical Computing, Vienna, Austria). P<0.05 was considered to indicate a statistically significant difference.

Of the 91 patients, 3 patients had a history of myelodysplastic syndrome. No patient had other hematological disorders at diagnosis. The median hemoglobin was 78.0 g/l (36.0–136.0 g/l), the median white blood cell and platelet count was 29.5×10⁹/l (1.1–417.8×10⁹/l) and 36.0×10⁹/l (4.0–305.0×10⁹/l), respectively. A total of 15 patients exhibited hyperleucocytosis (white blood cells>100×10⁹/l). The blast count in bone marrow aspiration at diagnosis ranged between 28.5 and 97.0%, the median blast count was 68.5%. The clinical and hematological characteristics of these patients are listed in Table I. It was observed that 10/73 (13.7%) patients exhibited FLT3/ITD, and FLT3/TKD was positive in another 2 patients. Karyotypic studies revealed unfavorable chromosomal abnormalities in 18 patients, including −7, 3q, 11q and complex karyotype. A total of 11 and 53 patients were in favorable and intermediate groups, respectively. The other 9 patients exhibited no assessed cytogenetic factors. The risk categories were defined according to the SWOG criteria. A total of 44 patients experienced primary induction failure and 47 cases experienced relapse, including 20 cases of early relapse (42.6%) and 27 cases of late relapse (57.4%).

All 91 patients were assessed until the final follow-up and 68/91 (74.7%) patients achieved CR (CR + CR with incomplete platelet recovery) following the MAC regimen. The CR rate was 72.7 (32/44) and 76.6% (36/47) in patients with primary induction failure and those who experienced relapse, respectively. For the relapsed cases, it was 65.0 (13/20) and 85.2% (23/27) in cases experienced early relapse and late relapse, respectively. No difference in the CR rate was observed between these groups. A total of 4/10 patients with FLT3/ITD positive achieved CR following the MAC regimen, PR and non-remission (NR) was observed in three cases, respectively. The overall CR rate was 81.8 (9/11) and 79.2% (42/53) for patients with favorable and intermediate cytogenetic groups, while it was 66.7% (12/18) for the unfavorable cytogenetic group. No difference in CR rate was observed for patients in different cytogenetic risk groups. The patients with unavailable cytogenetic factors had a CR rate of 55.5% (5/9). Early mortality was observed in only one case during the agranulocytosis period following chemotherapy, as a result of a severe lung infection and gastrointestinal hemorrhage.

A total of 27 patients underwent at least two high-dose cytarabine (Ara-c; 3 gm⁻² q12 h on days 1–3) or intermediate dose Ara-c plus anthracyclines prior to the MAC regimen. A total of 7/27 patients underwent SCT following the MAC salvage regimen.

Among the 68 patients who achieved CR following the MAC regimen, five underwent allogenic-SCT, one succumbed to relapse following SCT, and four remain in CR. The other 63 patients entered consolidation chemotherapy. Among them, 26 patients were referred to 1–3 courses high-dose Ara-c (3 gm⁻² q12 h on days 1–3) or intermediate dose Ara-c plus anthracyclines consolidation chemotherapy, other patients received consolidation chemotherapy with daunorubicin, homoharringtonine, mitozantrone or amsacrine combined with a sd of cytarabine. A total of 32 patients remain alive in CR, 2/31 patients with relapse succumbed to subsequent progressive disease, and 2 patients with partial remission following the MAC regimen achieved CR with a sd of mitoxantrone and cytarabine. Each patient underwent allogenic-SCT and remain in CR. Among the 21 patients with NR, 18 failed to achieve CR even though they underwent G-CSF (300 µg) administered 12 hours prior to fludarabine (30 mg⁄m²) in 0.5 h infusions for 5 consecutive days and cytarabine (2 g⁄m²) in 4 h infusions for 5 consecutive days (FLAG), aclacinomycin (5–7 mg⁄m²) for 8 consecutive days, cytarabine (10 mg⁄m²) for 14 consecutive days and G-CSF (300 µg) regimen re-induction and the final three were lost at follow-up.

The median follow-up duration was 46 months (range, 2–69 months). The median OS of the 91 patients was 22 months (range, 0.5–73 months; Fig. 1). The OS at 1, 3 and 5-years for the entire study population was 72.1, 42.9 and 36.7%, respectively (Fig. 1).

For patients who attained CR (n=68) or failed to (n=23), 1 and 3-year OS rate were 83.3, vs. 32.7% (P<0.001) and 57.3, vs. 10.3% (P<0.001), respectively (Fig. 2). The median survival was 12 (4–73) and 5 (0.5–20) months in patients with a white blood cell count <100×10⁹/l and ≥100×10⁹/l (P=0.0001), and it was 7.5 and 4 months in FLT3 positive and negative patients, respectively (P=0.0001). In a multivariate analysis, the white blood cell counts at diagnosis and FLT3 mutation were independent factors for OS (P=0.035 and P=0.008).

For the entire cohort, the 1 and 3-year DFS rates were 48.1 and 38.0%, respectively. The median DFS was 10 months (Fig. 3). For patients who attained CR following the MAC regimen, the DFS rate at 1, 3 and 5-years was 59.7, 47.1 and 43%, respectively (Fig. 4). For the OS rate (P=0.32), as well as the DFS rate (P=0.19), no difference was observed in different cytogenetic risk groups.

The median OS and DFS in seven transplanted patients was not significantly different compared with the group treated with chemotherapy alone, and they were 31 and 17, and 23 and 12 months, respectively (P=0.32 and P=0.18).

Hematological toxicity and infections were the most prominent toxicities of the MAC treatment. All patients developed grade IV neutropenia, according to the NCI CTC4.0 criteria following the MAC therapy. The median duration of neutrophils <0.5×10⁹/l and platelets <20×10⁹/l was 11 (range, 6–22) days and 9 (range, 5–25) days, respectively. The patients received a median of 10 units of packed red cells (range, 3–29) and 6 units (range, 2–12) of platelets. The median duration of hospitalization was 26 (range, 19–56) days. Severe infections (grade III or IV) were observed in 29% (n=26) of the cases, among which fungal was observed in 8% (n=7). The median duration of antibiotic therapy was 11 days. A total of 5 (5.5%) patients had lightly elevated levels of aminotransferase and nausea, which resolved shortly following the completion of chemotherapy. Liver, kidney, circulatory failure or neuropathy following the MAC regimen was not observed.