We are standing on the precipice of a new era in oncology drug development. In his Pulitzer prize winning book, The Emperor of All Maladies: A Biography of Cancer, Dr. Siddhartha Mukherjee traces the evolution of the approach to treating cancer. He identifies four major paradigms in our understanding of this disease:
- Ancient views of cancer: The oldest recorded description of cancer was found on a papyrus from about 1600 BCE. The ancient clinician described the probable case of breast cancer as “a bulging tumor . . . like touching a ball of wrappings.” As for treatment? The papyrus reads that there is none.
- The dawn of modern treatment: In 1947, Dr. Sidney Farber developed the first chemotherapy wherein he achieves a brief remission in a toddler suffering from acute lymphoblastic leukemia by treating him with aminopterin.
- If some is good, more is better: In an attempt to help patients, oncologists performed radical surgeries and gave massive doses of chemotherapy in combination with bone marrow transplants, to little avail.
- The quest for targeted treatments: Instead of simply killing cancerous and normal cells with toxic drugs, the newest generation of oncology drugs more selectively inhibits hyperactive growth pathways in cancer cells.
The emergence of pharmacometric analyses is the latest leap forward in how we bring safer, more effective medications to cancer patients. It can bring great value to sponsors by helping them optimize dosing, inform the drug label, achieve regulatory compliance, and in some cases, waive clinical trials.
Pharmacometric analyses encompass a number of methods including:
- PBPK modeling
- Population pharmacokinetic (pop PK) analysis
- Pharmacokinetic/pharmacodynamic (PK/PD) modeling
- Concentration/QTc analysis
- Exposure-response modeling
I will discuss how three of these methods have been invaluable to our clients’ oncology programs.
PBPK models for DDI and organ impairment
- Modeling objectives: The Simcyp Population-based Simulator streamlines drug development through the modeling and simulation of physiologically-based pharmacokinetics (PK) and pharmacodynamics (PD) in virtual patient populations. It incorporates numerous databases containing human physiological, genetic and epidemiological information. By integrating this information with in vitro or clinical data, the Simulator can predict PK/PD behavior in ‘real-world’ populations. It can be used to select ideal dosing regimens; determine drug-drug interactions (DDIs); and predict PK changes in special populations, such as pregnant women, children, or patients with organ impairment.
- How does it add value to your program? Cancer patients frequently take drug regimens that include multiple compounds. This means that they may face elevated risks for DDIs. PBPK models can facilitate designing DDI studies and predicting the magnitude of DDI in various clinical situations. This approach can also be used to assess the impact of renal and hepatic impairment on drug PK. In some cases, insights gleaned from PBPK models can even eliminate the need for dedicated phase I studies in healthy volunteers for drug interactions and organ impairment.
Population PK analysis in early clinical development
- Modeling objectives: Using intensive and sparse PK samples collected in phase 1/2 study in cancer patients as input data, population PK models can be built to assess covariate relationships between model parameters and patient demographics and baseline characteristics (body size metrics, age, sex, race and measures of renal and hepatic function). This model can then be leveraged to perform PK simulations to assess between-subject variability in exposures metrics (AUC, Cmax) when the investigational drug is given by weight-based and flat dose levels. Optimization software, such as PFIM or PopED can be used to develop a sparse sampling schedule by determining the optimal number and timing of PK samples.
- How does it add value to your program? This approach can aid justifying the dosing regimens for subsequent trials and employing sparse PK sampling in trials to decrease burden of collecting PK in patients at multiple centers.
Concentration-QTc analysis in phase 1/2 studies
- Modeling objectives: The input data for these analyses is time-matched ECG recordings and PK samples collected in the Dose Escalation and Expansion phases of phase 1/2 studies. Electrocardiogram (ECG) recordings should be collected using same rigor as those collected in TQT study. These data can be used to build a model that characterizes the relationship between plasma parent drug/metabolite concentrations and baseline-adjusted Fridericia-corrected QTc (QTcF) interval (∆QTcF). Improved understanding of the concentration-QTc relationship may inform whether the drug and/or metabolites prolongs the QTcF interval when administered as a single agent or in combination with other medications in phase 1/2 studies. It also might provide insight into which subject covariates cause the variability in ∆QTcF.
- How does it add value to your program? This approach can fulfill regulatory requirements of ICH E14 and waive the need for dedicated QT study in cancer patients. In addition, the results may be used for the Section 12.2 Pharmacodynamics of the product label.
use pharmacometrics to inform drug