High Burden of Healthcare Utilization in Older Acute Myeloid Leukemia Patients Treated with First-Line Venetoclax-Based Therapies

BACKGROUND: Venetoclax with azacitidine (VEN/AZA) or low-dose cytarabine (VEN/LDAC) is the standard of care for acute myeloid leukemia (AML) patients who are unfit for intensive chemotherapy. However, there is limited insight into the real-world treatment delivery rate and healthcare burden.

METHOD: This retrospective multi-institutional study in Australia analyzed AML patients treated with first-line VEN/AZA or VEN/LDAC between January 2021 and May 2024 to evaluate treatment delivery rate and healthcare utilization metrics, such as the frequency and duration of unplanned hospitalizations and red blood cell (RBC) and platelet transfusions. We also assessed the impact of patient-related and disease parameters on treatment delivery, healthcare utilization and outcome.

RESULTS: Of the 102 AML cases, 96 (94.1%) received first-line VEN/AZA and 6 (5.9%) received VEN/LDAC treatment. The median age at diagnosis was 74 years (interquartile range [IQR] 67.9, 79.0).

The majority of the evaluable patients (n=92; 90.2%) had intermediate-risk (n=55; 59.8%), followed by adverse-risk (n=39; 42.4%) cytogenetics. However, according to the 2022 European LeukemiaNet prognostic system (ELN 2022), most of the evaluable patients (n=79; 77.5%) were classified as adverse risk (n=58; 73.4%), followed by intermediate (n=13; 16.5%) and favorable-risk (n=6; 7.5%).

The median number of treatment cycles delivered was three (IQR 1.3, 6). Of the 82 patients who had response assessment after first cycle, 61% (n=50), 9.76% (n=8), 12.2% (n=10) and 17.1% (n=14) achieved complete remission with or without complete count recovery (CR/CRi), morphologic leukemia-free state (MLFS), partial remission, and refractory disease, respectively.

The majority of evaluable patients (n=48) received antifungal prophylaxis (n=34; 70.8%), the most common being posaconazole (n=33) requiring venetoclax dose reduction. Serum posaconazole levels ≥0.7 mg/L were maintained in 86.2% of the 282 samples tested. In the first two cycles, the evaluable patients treated with posaconazole (n=33) received a lower venetoclax dose than patients treated without posaconazole (n=15) (102±104 vs. 167±183 mg) and (82.7±69.2 mg vs. 119±194 mg), with similar dosing from the sixth cycle onwards (56.3±17.1 mg vs. 50±0 mg).

The mean duration of venetoclax treatment during the first five cycles was 25.9±4.9, 23.8±6.9 and 23.2±6.2, 20.5±6.6, 18.7±6.9 days, respectively, which was reduced to 17.3±7.2 days in the sixth cycle. Despite venetoclax dose adjustment, 59% and 27.6% of treatment cycles were delayed to 35 days and 42 days, respectively.

At the time of last follow-up, 88.2% (n=90) of patients discontinued therapy, including 53.9% (n=55) who died. The median OS of the whole cohort was 13.6 months, with a significantly longer OS observed in patients who achieved CR/CRi (13.2 vs. 5.9 months; P <0.001). The median OS of TP53-mutated patients was significantly poorer than those with wild-type TP53 (4.4 vs. 12.03 months; P=0.001).

During the treatment course, 96 of the 102 patients (94%) required 390 hospitalizations for a total of 4193 days. In the first cycle, 88 (86.3%) patients required a total of 106 hospitalizations for a total of 391 days. 71% (n=32) of the evaluable patients had age adjusted Charlson Comorbidity Index (CCI) of ≥4. We did not observe a difference in the mean number (5.7 vs. 6.3, P=0.9) and duration of hospitalisations (65.4 vs. 51.2, P=0.6). Similarly, the mean number of cycles completed between the two groups was comparable (5.6 vs. 8.0, P=0.5) between patients with a CCI of 1-3 or ≥4.

All evaluable patients (n=48) required RBC transfusions, with a median of 17.0 units (IQR 8.0, 38.5) per patient. During cycles 1 and 2, 12.5% and 8.8% of patients required RBCs, which decreased to 7.7% in cycle 3. Among the 16 evaluable patients who received ≥1 phenotype-mismatched RBC unit, 37.5% (n=3) developed RBC alloantibodies (anti-D, C, Jkb, Fya). Platelet transfusions were required by 93.3% (n=42) of patients, with a median of 9.50 (IQR 3.00, 29.3) units per patient.

CONCLUSION: VEN/AZA and VEN/LDAC therapies are associated with improved survival but require substantial healthcare support, including recurrent prolonged hospitalization and transfusion support. In addition, despite dose adjustment, ~60% of cycles could not be delivered to the 28-day schedule, with 27% delayed ≥42 days.