Gentuzumab-Ozogamicin,Citosine Arabinoside, G-CSF Combination in the Treatment of Elderly Poor Prognosis Acute Myeloid Leukemia. A Multicentric Study.

Gentuzumab Ozogamicin (GO) is an anti-CD33 antibody linked to the cytotoxic antibiotic calicheamycin and it is effective as single agent in the treatment of poor risk acute myeloid leukemia (AML) patients. In 3 Italian Hematology Departments from September 2003 to December 2004, we treated 26 AML patients, both untreated (12 cases) and resistant (14 cases) with the following protocol: rhG-CSF(5 μg/kg, on days 0 through 8), Aracytin as continuous perfusion (100 mg/m² on days 4 through 8) followed by GO (6 mg/m² iv over 2 hours on day 9) (G-GOA). Inclusion criteria were: 1) CD33+ de novo or secondary AML (except M3 AML; 2) Primary refractory AML or relapse of AML in patients between 61 and 80 years; 3) Untreated patients 〉70 years or not eligible for aggressive chemotherapy. There were 13 male and 13 female with a median age of 69 years (range 58–77). FAB classification was 5 M0, 5 M1, 7 M2, 2 M4, 1 M5, 6 AML post-MDS. Ten patients presented a secondary AML. The median duration of first complete remission (CR) of 9 patients with relapsed AML was 48 weeks (range 8–76). Cytogenetic study was performed in all patients; karyotype was intermediate in 13 cases, unfavourable in 7 cases. In 6 patients no metaphases were observed. All patients performed the CD33+ evaluation on BM, the median percentage of CD33 positive blasts was 90% (range 25%–95%). Fourteen patients (56%) achieved a CR and 1 patient had CR with delayed platelet recovery (CRp) with an overall response (OR) of 58% (7 untreated AML and 8 resistant patients). One patient obtained a partial remission with only a transient hematologic improvement, characterized by a peripheral increase of all hematological parameters and by a 50% reduction of the bone marrow blast count. Five patients (19%) resulted refractory to treatment and 5 patients died during the aplasia period post induction treatment. The most common adverse event was myelosuppression, as expected. Median durations of neutropenia and thrombocytopenia in patients reaching CR were 19 days (range 15–62) and 16 days (range 10–37) respectively. No VOD was recorded. Six patients (23%) developed documented infection (including pulmonary aspergillosis in 3 cases). Notably, in 2 cases we observed a grade III/IV bleeding consisting in CNS haemorrhage. Two patients died while in CR, 1 due to bladder cancer relapse and 1 to ischemic stroke. Seven patients (27%) relapsed; the median CR duration was 20 weeks (range 8–41+). At March 2005 10 patients (38%) were alive of whom 6 are still in CR (27%). The median overall survival was 17,5 weeks (1–60 weeks). On the basis of our results the G-GOA protocol could be considered an useful approach for poor risk elderly AML also in consideration of the low reported side effects. We did not observe any case of VOD, which characterized other trials using GO at higher dosage. Unfortunately the CR duration was brief. Probably it could be improved with the addition of more aggressive consolidation schedule (i.e. Intermediate doses of Ara-C).