043: Semi-Mechanistic Model of Tumor Lysis Syndrome in Chronic Lymphocytic Leukemia Patients

Statement of Purpose: Venetoclax is a first-in-class BCL2 inhibitor approved for treatment of chronic lymphocytic leukemia (CLL). In this study, we developed a semi-mechanistic pharmacokinetic/pharmacodynamic model to investigate the relationship between venetoclax pharmacokinetics and potassium excursion events among CLL patients with low/medium tumor lysis syndrome (TLS) risk.

Description of Methods & Materials: The semi-mechanistic model included three major components: venetoclax pharmacokinetics, lymphocyte growth and death kinetics in blood as well as potassium dynamics (Figure 1). Venetoclax effect on the death rate of lymphocytes was modeled as an Emax model. Intracellular and extracellular potassium levels were modeled as separate compartments to describe potassium dynamics. Extracellular potassium clearance was modeled proportional to baseline creatinine clearance and via an aldosterone-dependent feedback mechanism. This model was fit to baseline creatinine clearance and longitudinal concentrations of potassium in plasma from 6 venetoclax CLL studies that included 538 front-line or relapsed/refractory patients. NONMEM® [v7.5.0] with the SAEM optimization algorithm was used to estimate parameters and inter-individual variabilities.

Data & Results: All the estimated population parameters were physiologically reasonable (Table 1). The semi-mechanistic model adequately described the observed potassium excursions (defined as >6 mEq/L by Howard criteria) in low/medium TLS risk subjects across most studies, with predicted medians close to observed values in R/R studies. However, the model overestimated potassium related TLS events in the front-line study due to the difficulty in capturing rare events in model predictions and the low incidence of observed potassium lab TLS events (N=2). When the threshold for potassium excursion was changed from 6 mEq/L to 5 mEq/L, the model accurately described the more frequently occurring excursions for the front-line study.

Interpretation, Conclusion or Significance: These analyses represent the first population semi-mechanistic approach to describe potassium excursions in low/medium TLS risk CLL subjects. The model can be leveraged to predict potassium related TLS rate for alternative study designs or dosage regimens. It can be further extended to include additional laboratory parameters namely, phosphorus, calcium and uric acid to understand if the association between these electrolytes and venetoclax exposure is capable of accurately predicting TLS events across all patient populations.

Disclosures: This study was supported by AbbVie in collaboration with Genentech/Roche. AbbVie and Genentech provided financial support for the study and participated in the design, study conduct, analysis, and interpretation of data as well as the writing, review, and approval of the manuscript. NT, IW, BN, SS, JW, SM, RM and AHS are employees of AbbVie and may hold stock. WH is a Genentech employee and may own stock.