Gemtuzumab ozogamicin for salvage CBF-AML: A case report

During the 2^nd How to Diagnose and Treat Acute Leukaemia conference organized by the European School of Haematology (ESH), there was an interactive session dedicated to discussing the off-label use of mutation-based therapy in AML. Nicolas Boissel presented the available data from gemtuzumab ozogamicin (GO) administration in a particular patient with core binding factor (CBF)-AML.¹

In 2018, GO was approved for use in addition to standard induction chemotherapy in newly diagnosed patients with CD33⁺ AML. This anti-CD33 conjugated monoclonal antibody had been previously authorized in 2000 through an accelerated review process, but was withdrawn in 2010 from the market due to increased mortality rates. With the current indication in a specific population and modified dosing schedule, GO’s safety profile improved, and it is being further investigated in several studies as frontline and salvage therapy for AML (e.g., the AML18 trial).

Patient details¹

• 35-year old male
• Diagnosed with CBF-AML, M2 classification
  • t(8;21) – CBF subunit A (CBFA)
  • KIT mutated (variant allele frequency [VAF], 20%)

Treatment¹

Frontline treatment

• Induction regimen: 7+3 with cytarabine 200 mg/m² and daunorubicin 90 mg/m², followed by three courses with cytarabine 3,000 mg/m² every 12 hours on Days 1, 3, and 5 as consolidation
• The patient showed a rapid and profound response, with a more than 3-log reduction in measurable residual disease before the second cycle of treatment
• Hematopoietic stem cell transplant (HSCT) was saved for salvage therapy, since initial response was satisfactory
• 15 months after the first reported complete response, the patient experienced a hematological relapse; 3 months earlier, molecular relapse was detected
• Cytogenetic analysis was repeated, and persistence of KIT mutation was found (VAF 47%)

Treatment at relapse¹

• Data on relapsed CBF-AML is scarce and generally from retrospective studies (Table 1)
• CBF-AML complete responses at relapse are 65–85% when treated intensively, and 5-year OS is 32–51%. Overall, CBFA presents worse survival outcomes than CBFB
• The most important clinical prognostic factor in relapsed CBF-AML is the duration of the first complete response. The role of allo-SCT at second complete response is controversial when comparing different published studies (Table 1)

Table 1. Summary of retrospective studies reporting the outcomes of patients with relapsed CBF-AML¹

GO for CBF-AML

• The efficacy of GO has been primarily investigated in first-line randomized studies, and mostly observed in patients presenting a favorable karyotype¹
• In a recent French study, the ALFA-0701 trial (NCT00927498), the investigators reported that patients showing the most benefit from first-line GO (more prolonged event-free survival), were those harboring signaling mutations, which are associated with CD33 overexpression⁶
  • Of note, patients with CBF-AML were excluded from this posterior analysis. However, as previously published, GO efficacy is higher in those patients with a higher expression of its target, CD33
• The benefit of GO in relapsed CBF-AML, in terms of OS, was first reported in a retrospective study by the French AML Intergroup, with 78 patients who received allogeneic HSCT in second complete response. They observed that those who received GO before transplant significantly improved post-transplant survival outcomes, without significant increment of treatment-related mortality⁷

Conclusion¹

• Patients with CBF-AML presenting signaling mutations might benefit from GO in first-line and relapse
• Additional studies are needed to confirm if the use of GO in CBF-AML should be restricted to patients with signaling mutations, which are frequently reported in this AML subtype
• The efficacy of GO in this setting could be enhanced by combining it with targeted inhibitors
• Non-CBF-AML might also respond better to GO when signaling mutations are detected

At the Acute Leukemias XVII Biology and Treatment Strategies biennial international symposium in 2019, the AML Hub talked with Alan K. Burnett from Glasgow University, Scotland, UK, about GO’s role in CBF-AML. See below his key messages on this topic: