Patient outcomes in AML: Comparison between randomized clinical trials and registry data

The strict eligibility criteria for clinical trials in acute myeloid leukemia (AML) often result in a lack of generalizability to real-world patient outcomes.¹ In addition, accounting for geographical differences in treatment outcomes can be challenging, and the wide variation in clinical practice has become more complex, particularly regarding the introduction of novel first-line therapies.¹

Recently, Tiong et al.¹ published a comparative analysis in Blood Cancer that highlights the similarities and differences in the standard of care in clinical trials vs clinical practice across Australia. We summarize the key findings below.

Methods¹

• Patients were enrolled in the Australasian Leukemia and Lymphoma Group National Blood Cancer Registry (ALLG NBCR) between December 2012 and April 2018.
• Matched controls from published randomized trials that included first-line AML therapies were used. 

Key findings¹

• A total of 942 patients were enrolled in the ALLG NBCR cohort.
• Idarubicin was the most common anthracycline received by patients in the ALLG NBCR cohort.
• In the ALLG NBCR cohort, induction regimens involved higher doses of cytarabine (1,000–3,000 mg/m²) compared with clinical trials, where 60 mg/m² daunorubicin and 100–200 mg/m² cytarabine were used.

Patients with FLT3-mutated AML aged 18–59 years (ALLG NBCR vs RATIFY [NCT00651261])

• A higher proportion of patients in the ALLG NBCR cohort underwent hematopoietic stem cell transplantation (HSCT) in first remission compared with those in the RATIFY trial (46% vs 25%; p < 0.001).
• Median overall survival (OS) was 45.7 months in the ALLG NBCR cohort vs 25.6 and 74.7 months, respectively, in the placebo and midostaurin arms of the RATIFY trial.

Patients with de novo AML aged 50–70 years (ALLG NBCR vs ALFA-0701 [NCT00927498])

• A greater proportion of patients in the ALLG NBCR cohort received an allogeneic HSCT in first remission compared with those in ALFA-0701 trial (21% vs 14%; p = 0.03).
• 2-year event-free survival (43% vs 17%) and 2-year OS (54% vs 42%) were higher in the ALLG NBCR cohort than the ALFA-0701 trial. 

Patients with secondary/therapy-related AML aged 60–75 years (ALLG NBCR vs CPX-351 trial [NCT01696084])

• Rates of CR or CR with incomplete count recovery (CRi) were higher in the ALLG NBCR cohort (53% vs 33%; p = 0.005) and median OS was longer (9.5 vs 5.95 months) vs the 7 + 3 induction arm of the CPX-351 trial.
• More patients in the CPX-351 trial received allogeneic HSCT than in the ALLG NBCR cohort (29% vs 7%; p < 0.001).

Patients aged ≥55 years in first remission after intensive chemotherapy (prior to the approval of oral azacitidine) (ALLG NBCR vs QUAZAR AML-001 [NCT01757535])

• Patients in the QUAZAR AML-001 trial who received 0, 1, or ≥2 consolidation cycles had CR/CRi rates of 20%, 45%, and 35%, respectively, compared with 11%, 70%, and 18% of patients in ALLG NBCR cohort (p < 0.001).
• Patients in the ALLG NBCR cohort who did not undergo allogeneic HSCT had a median OS of 32.2 months vs 14.8 and 24.7 months in patients receiving placebo or oral azacitidine, respectively, in the QUAZAR AML-001 trial. 

Key learnings

This comparative analysis revealed differences in patient outcomes in the context of Australian clinical practice prior to the introduction of novel first-line therapies. Challenges of evaluating the comparative benefit of novel therapies in patients with AML using real-world data were highlighted, which are further exacerbated by the wide variation in global clinical practice.