Abstract
Venetoclax has become a critical component of frontline therapy in acute myeloid leukemia (AML) and an emerging option in selected multiple myeloma (MM) populations, though its toxicity profile across these distinct diseases remains poorly characterized. In AML, venetoclax is typically combined with azacitidine for older or unfit patients, while in MM, it is most commonly used in patients with t(11;14), paired with bortezomib. Despite these differing clinical contexts, venetoclax's central role in both regimens presents a unique opportunity to compare early toxicity signals across hematologic malignancies.
We hypothesized that early acute kidney injury (AKI) and transfusion need may serve as real-world biomarkers of treatment intolerance in AML, reflecting a higher regimen-specific vulnerability compared to MM.
Using the TriNetX US Collaborative Network, a federated real-world database encompassing 69 healthcare systems, we retrospectively analyzed adults with newly diagnosed AML or MM who received venetoclax-based therapy. Patients with chronic kidney disease or end-stage renal disease were excluded to isolate incident AKI. Outcomes were assessed beginning one day after therapy initiation and interpreted as occurring during the early treatment phase. The primary endpoint was incident AKI (ICD-10 N17). Transfusion (CPT 36430) was evaluated as a surrogate for early marrow suppression. Mortality and ICU admission were analyzed for contextual interpretation. No propensity score matching or multivariable adjustment was applied.
A total of 3,806 AML and 5,673 MM patients met inclusion criteria. AKI occurred in 29.5% of AML patients versus 20.3% of MM patients (risk ratio [RR] 1.45; 95% confidence interval [CI] 1.35–1.56; odds ratio [OR] 1.64; 95% CI 1.49–1.81; p<0.001). Transfusion occurred in 72.6% of AML patients and 27.4% of those with MM (RR 2.65; 95% CI 2.53–2.77; OR 7.00; 95% CI 6.39–7.68; p<0.001). Mortality was higher in AML (61.3%) than MM (28.8%) (RR 2.13; p<0.001), consistent with underlying disease trajectory. ICU admissions were rare (0.3% in both groups) and excluded from further analysis.
This comparative analysis revealed a substantially higher incidence of early AKI and transfusion among AML patients treated with venetoclax-azacitidine, underscoring a pronounced early toxicity burden. While some of this may reflect standard AML supportive practices, the frequency and early timing of these complications suggest a potential regimen-specific vulnerability—possibly due to venetoclax-induced tumor lysis, azacitidine-driven cytopenias, or reduced physiologic reserve in this population. These differences are unlikely to be explained solely by disease aggressiveness. AML patients are often older, frailer, and more heavily cytopenic at baseline, which may compound the toxicity profile of venetoclax-azacitidine. Additionally, rapid tumor debulking and cell death associated with venetoclax could exacerbate renal vulnerability, particularly in the setting of azacitidine-induced marrow suppression. In contrast, MM patients are more likely to receive therapy in the outpatient setting with lower-intensity regimens and longer pre-treatment stabilization, which may partially mitigate early organ stress. Monitoring early toxicity signals in real time may enable clinicians to personalize dosing, adapt schedules, or initiate preemptive supportive interventions—ultimately improving regimen completion and clinical outcomes.
Although tumor lysis labs, performance status, dose variation, and inpatient versus outpatient care setting were not available in this dataset, these unmeasured confounders may have contributed to observed toxicity patterns. Nevertheless, these findings support the hypothesis that early AKI and transfusion could serve as practical toxicity biomarkers to identify AML patients at risk for poor treatment tolerance. If prospectively validated, these markers may inform risk-adapted dosing, early triage to supportive interventions, or improved clinical monitoring in real-world AML care.
