Phase II–III Study of Gemtuzumab Ozogamicin Monotherapy versus Best Supportive Care in Older Patients with Newly Diagnosed AML Unfit for Intensive Chemotherapy: First Results of the EORTC-GIMEMA AML-19 Trial

Treatment of AML in the elderly remains challenging. In particular, few therapeutic options exist for patients (pts) who are not considered medically fit for intensive chemotherapy. Gemtuzumab ozogamicin (GO), a humanized anti-CD33 monoclonal antibody conjugated to calicheamicin, has demonstrated single agent efficacy in older patients with relapsed AML. In a previous trial we reported a complete response rate of 17% when the licensed dose/schedule of GO was used as frontline monotherapy for older unfit pts with AML, but excess hematologic and liver toxicity suggested changes in both dosing and scheduling to improve feasibility. The AML-19 study was designed as a sequential phase II–III trial in pts of age 61 or over with untreated AML, who could not be considered candidates for, or must have declined, conventional intensive chemotherapy. The primary objective of the phase II study was to investigate whether any of two different schedules of lower dose GO monotherapy (total delivered dose 9 mg/m² in 2 or 3 fractions over one week) is sufficiently effective and tolerable frontline therapy to continue phase III comparison with best supportive care (BSC). Pts had to have adequate renal and hepatic function, and the WBC count had to be less than 30,000/cmm at the time of registration. Two-thirds of the pts were randomly assigned to receive a single induction course of either GO 3 mg/m² iv on days 1, 3 and 5 (arm A), or GO 6 mg/m² iv on day 1 and 3 mg/m² on day 8 (arm B); the remaining third was assigned to the BSC arm. Randomization was stratified by age, performance status (PS), initial WBC, CD33 expression on marrow blasts, and Institution. Pts with no evidence of objective disease progression were eligible to receive continuation treatment with monthly outpatient infusions of GO at 2 mg/m² for a maximum of 8 months. Primary end-point of the phase II study was clinical benefit rate (CBR) defined as the number of pts either achieving a remission (CR, CRp, PR) or maintaining a stable disease (SD) in each experimental arm. The arm associated with the highest CBR would be selected for phase III assessment, providing that it is at least 48% (Simon design). Between 06/2004 and 12/2006, 56 pts were randomized in the two experimental arms (arm A 29; arm B 27). The two arms were comparable in terms of age (arm A: median 77 yrs; arm B: median 78 yrs), PS >2 (arm A: 7%; arm B: 4%), secondary AML (arm A: 45%; arm B: 37%), and CD33 expression on ≥20% marrow blasts (arm A: 86%; arm B: 85%). In arm A, 6 pts achieved CR, 2 PR, and 3 maintained SD for an overall CBR of 38% (90% CI, 23%–55%); in arm B, 5 pts achieved CR, 1 CRp, and 11 maintained SD for an overall CBR of 63% (90% CI, 45%–78%). Most toxicities were <grade 3 in both arms and included nausea/vomiting, diarrhea, fatigue, stomatitis and transient elevations of serum transaminases and bilirubin. The rate of grade ≥3 infection was slightly higher in arm B (48% vs 31%), as was the incidence of severe bleeding (11% vs 0%). All-cause early mortality (<6 wks) occurred in 17% of pts in arm A (2 infection, 3 AML) and in 11% of pts in arm B (1 infection, 1 AML, 1 infection+AML). In summary, lower dose GO is a tolerable and active primary therapy for older AML pts who are considered unsuitable for intensive chemotherapy. The day 1+8 schedule (arm B) is associated with a more favourable efficacy profile, meeting the statistical criteria to be selected as the preferred regimen for phase III comparison with BSC. Accrual to the phase III trial is ongoing.