Replacing the Anthracycline By Gemtuzumab Ozogamicin in Older Patients with De Novo Standard-Risk Acute Myeloid Leukemia Treated Intensively - Results of the Randomized ALFA1401-Mylofrance 4 Study

Introduction

Based on the results of the ALFA-0701 trial (Castaigne et al. Lancet 2012), the addition of fractionated doses of the anti-CD33 antibody drug conjugate gemtuzumab ozogamicin (GO, Mylotarg®) to conventional chemotherapy has been approved in 2017 for frontline treatment of adults with CD33-positive acute myeloid leukemia (AML). Prolonged event-free survival (EFS) was observed in patients with AML of favorable or intermediate risk, while not in those with adverse cytogenetics. Nevertheless, more frequent added toxicities could make the addition of GO a questionable option in patients over 60-65 years of age. In this ALFA-1401/Mylofrance 4 trial (NCT02473146), we investigated if the replacement of the anthracycline by GO might also improve EFS in older patients.

Methods

Between January 2016 and March 2019, 225 patients were randomized 2:1 to receive an experimental GO-cytarabine combination (154 patients) or a standard anthracycline-cytarabine treatment (71 patients). Patients aged 65 to 80 years old (later extended to patients aged 60-64 years old), with previously untreated de novo AML of favorable or intermediate cytogenetics were eligible for the trial. Standard treatment arm consisted in a 7+3 using idarubicin at 12 mg/m ²/d on day 1 to 3 and cytarabine 200 mg/m ²/d on day 1 to 7. Experimental arm (GO arm) consisted of two doses of GO 3 mg/m ²/d on day 1 and 4 and cytarabine 200 mg/m ²/d on day 1 to 7. Post-remission therapy comprised two courses of intermediate-dose cytarabine (IDAC) at 1.5 g/m ²/12h on day 1, 3 and 5. In the GO arm, a third dose of 3mg/m ²/d GO was administered on day 1 of the first IDAC course. The first IDAC course could serve as second induction in patients not responding to the first one. The decision to perform allogeneic hematopoietic stem cell transplantation (allo-HSCT) in first remission was left to the discretion of the physician. The efficacy analysis was conducted in the modified intent-to-treat (mITT) population excluding patients who did not meet the cytogenetic eligibility criteria. The primary study endpoint was EFS. Secondary endpoints were response rate defined by complete remission (CR), CR with incomplete platelets recovery (CRp) and CR with incomplete hematological recovery (CRi), early mortality, relapse incidence, overall survival (OS) and safety.

Results

Among the 225 randomized patients, 214 (71 standard arm, 143 GO arm) were included in the mITT population. There were 126 men and 88 women. Median age was 70 years (61-80). Cytogenetics was of favorable and intermediate risk in 14 and 200 patients, respectively. One hundred and eighty-one patients reached CR/CRp/CRi, 64 (90%) in standard arm and 117 (82%) in GO arm (p=0.17). Median follow-up time was 38 months. At 2 years, estimated EFS was 38% [95% CI, 28-51] in the standard arm vs. 29% [22-37] in the GO arm (Hazard Ratio (HR), 1.37 [0.98-1.93]; p= 0.067) (Figure 1A). Overall, 118 patients relapsed, 36 (51%) in standard arm and 82 (57%) in GO arm. At 2 years, estimated cumulative incidence of relapse was 48% [36-61] in standard arm vs. 61% [52-70] in GO arm (csHR, 1.32 [0.90-1.93]; p= 0.078). Overall, 122 patients died, 38 (54%) in standard arm, 84 (59%) in GO arm. Sixty-day mortality was 4% in standard arm vs. 10% in GO arm (p=0.13). At 2 years, estimated OS was 65% [55-77] in standard arm vs. 52% [45-61] in GO arm (HR, 1.27 [0.86-1.87]; p= 0.23). In subgroup analysis for EFS (Figure 1B), we found a significant interaction with gender, GO having a detrimental effect in women which persisted after adjustment on known prognostic factors. A total of 33 patients received allo-HSCT in first remission, 19 (30%) in standard arm and 14 (12%) in GO arm (p=0.006). When censoring these patients at transplant time, HR of GO on EFS was 1.27 ([0.88-1.83]; p=0.19). Regarding safety, 76% and 80% of patients had at least one grade 3 to 5 adverse event (p=0.81), including infection in 30% vs. 21% and bleeding in 7% vs. 29% in standard arm and GO arm respectively. Serious adverse events were reported in 34% of patients in standard arm vs. 49% in GO arm (p=0.031). Sinusoidal obstruction syndrome occurred in 2 patients in the GO arm.

Conclusion

Frontline use of GO instead of idarubicin, when combined to cytarabine, does not benefit older patients with de novo standard-risk AML. At the reduced dose schedule used in this study, GO remains associated with significant toxicities while non-significant higher relapse incidence, shorter EFS and shorter OS were observed.

Figure 1

View largeDownload slide

Figure 1

View largeDownload slide

Close modal