The Special Opportunities for Modeling and Simulation in Oncology Drug Development

In his most recent New York Times Magazine piece, “The Improvisational Oncologist,” Dr. Siddhartha Mukherjee, author of The Emperor of All Maladies: A Biography of Cancer, wrote, “In an era of rapidly proliferating, precisely targeted treatments, every cancer case has to be played by ear.”

Oncology treatment: yesterday, today and tomorrow

Mukherjee begins his article by describing his work in oncology, just a few years ago:

“I grew up as an oncologist in an era of standardized protocols. Cancers were lumped into categories based on their anatomical site of origin (breast cancer, lung cancer, lymphoma, leukemia), and chemotherapy treatment, often a combination of toxic drugs, was dictated by those anatomical classifications. The combinations—Adriamycin, bleomycin, vinblastine and dacarbazine, for instance, to treat Hodgkin’s disease—were rarely changed for individual patients. The prospect of personalizing therapy was frowned upon: The more you departed from the standard, the theory ran, the more likely the patient would end up being undertreated or improperly managed, risking recurrence. In hospitals and clinics, computerized systems were set up to monitor an oncologist’s compliance with standard therapy. If you chose to make an exception for a particular patient, you had to justify the choice with an adequate excuse. Big Chemo was watching you.”

And his perspective today:

“More and more, we must come up with ways to use drugs as precision tools to jam cogs and turn off selective switches in particular cancer cells. Trained to follow rules, oncologists are now being asked to reinvent them. Cancer, by contrast, has potentially unlimited variations. Like faces, like fingerprints—like selves—every cancer is characterized by its distinctive marks: a set of individual scars stamped on an individual genome. The iconic illness of the 20th century seems to reflect our culture’s obsession with individuality.”

From, the American Society of Clinical Oncology (ASCO) website:

Before personalized medicine, most patients with a specific type and stage of cancer received the same treatment. However, it became clear that some treatments worked better for some patients, than for others. The growth in the field of genetics has led researchers to find genetic differences in people and their tumors. In turn, this explained many of the different responses to treatment. A person with cancer may now still receive a standard treatment plan, such as surgery to remove a tumor. However, the doctor may also be able to recommend some type of personalized cancer treatment. Personalized cancer treatment is now an active part of the treatment plan or as part of clinical trial.

Clearly, personalized medicine is an evolving approach to cancer treatment. And while the Pharmaceutical Research and Manufacturers of America (PhRMA) report in the “Value of Personalized Medicine” that 73% of cancer medicines have the potential to be personalized medicines, we still don’t have all of the genetic information and experience to deliver to all patients. But, we are on the road.

The march toward individualized cancer treatment will leverage model-informed drug development

Oncology drug development is quite challenging, not just from a R&D perspective, but also an ethical one. Patient recruitment and retention, individualized variable response to treatment, and the ‘un-testability’ factors for fragile oncology patients pose unique barriers. Dosing, dosing regimen and formulation options in oncology are more complex as are the potential for clinically-significant drug-drug interactions (DDIs). Combinations of oncology drugs are often prescribed, adding additional layers of analysis for both efficacy and safety. Each of these variables can theoretically be ‘individualized’ for each patient and adjusted during the treatment period.

Model-informed drug development has demonstrated tremendous success over the past decade and is now routinely used and expected by global regulators. This is especially true for oncology where 100% of new drugs approved by the FDA in 2015 incorporated modeling and simulation.

Oncology drugs agents tend to have narrow therapeutic indices (TI), requiring precise dosing to ensure sufficient exposure for clinical activity while minimizing toxicity. These agents frequently have complex pharmacology, and combination therapy may cause schedule-specific effects and interactions. A wide variety of models are routinely developed during oncology drug development, including pharmacokinetic (PK), disease progression, and exposure-response modeling. Exposure-response modeling can include time-to-event (TTE) models and models relating exposure to adverse events (AEs). Physiologically-based pharmacokinetic (PBPK) models are applied to investigate DDIs and can be used to support dose and formulation recommendations. Modeling and simulation (M&S) helps identify safe starting doses for combination therapies as the clinical testing of all combinations of drugs and doses is not feasible. Models can also be used to determine informative times to assess exposure or response.

A major emerging contribution of M&S is the use of model-based estimates of tumor growth inhibition using longitudinal tumor size data. We can build models linking tumor growth inhibition to overall survival using historical data and those models can be leveraged to predict outcomes based on early clinical studies (Phase 1 or Phase 2) with new investigational treatments including combinations. Continued emphasis is being placed on understanding tumor size dynamics and the effects of investigational drugs on survival.

How is M&S used in oncology

The opportunities to leverage M&S to answer crucial questions span the entire oncology drug development cycle. In the translational stage, from pre-clinical to early clinical, we use these approaches to ask about optimal first-in-human (FIH) dosing, concomitant medications, and pharmacodynamic (PD) endpoints to help guide dose schedule options. In early clinical development, we can assess dosing and dose schedules with more precision, study formulation options, assess potential DDIs, and study more cohort- or patient-specific factors. In the late clinical stage, we use M&S to select the pivotal clinical trial dose that will provide optimum risk-benefit, perform bridging studies, and further study AE potential.

Specific opportunities for M&S are as follows:

  • Predict and characterize PK

Semi-physiological approaches can be used to predict PK at the site of action, in other populations.

Physiologically-based PK modeling is a powerful tool to understand the exposure at the site of action, i.e. the tumor. The comprehensive summary of the ‘system knowledge’ enables extrapolations between species (eg, from mouse to man) or between populations (eg, between adult and pediatric patients).

  • Characterize biomarker response in early clinical studies

Help define biologically effective doses – determine dose range for further clinical testing.

When an early clinical efficacy biomarker is available, PK/PD models for that biomarker allow the characterization of target engagement and help establish the dose regimens associated with pharmacological activity. Modeling provides an alternative to the maximum tolerated dose (MTD) paradigm that historically has been used in dose setting in oncology.

  • Translate pre-clinical data

Use mouse xenograft data to (further) support clinical dose regimen setting.

Especially in immuno-oncology, good clinical biomarkers are still lacking. A possible alternative to support the early clinical dose setting is the translation of pre-clinical efficacy data. For example, in the clinical development of the anti-PD1 antibody, the selection of the lowest and, ultimately, approved dose was largely built on a translational model framework through which mouse xenograft data were leveraged to predict clinically efficacious dose regimens.

  • Assess drug-drug and drug-food interactions

The sheer number of potential drug-drug interactions make clinical testing of all potential scenarios impossible and impractical, pointing to PBPK as an effective approach.

Through application of PBPK, a prospective DDI risk management strategy can be developed. PBPK models created in pre-clinical or early clinical development should be continuously updated with clinical PK data and mechanistic information on the molecule. PBPK has now been accepted by FDA to support 100 label claims, mainly in lieu of specific DDI studies.

  • Measure downstream biochemical or cellular effects of target/pathway modulation

Quantitative systems pharmacology is an emerging technology that sits at the interface between pharmacometric modeling and simulation and systems biology.

QSP allows prediction of the effects of multiple therapeutic interventions in combination. QSP can provide a framework to evaluate these potential combinations prior to clinical testing, by providing a quantitative understanding of how different mechanisms will interact.

  • Characterize variation in drug exposure (intrinsic/extrinsic factors)

In the absence of dedicated clinical pharmacology studies, population PK analysis of sparsely sampled patient data

In recent oncology drug approvals, the characterization of the pharmacokinetics and the impact of demographic and disease factors was based solely on integrated population PK analyses across the available patient data, rather than through dedicated phase 1 studies. As a result, all label statements around pharmacokinetics for these drugs are purely model-based.

  • Characterize tumor size responses

Help establishing optimal dose regimen and therapeutic window and enable use as early marker for survival (e.g. as tool in patient stratification).

Recently, several promising oncology drug products received their initial approval on the basis of (tumor size based) objective response data, rather than survival outcomes. This puts further emphasis on a thorough understanding of tumor size dynamics and the effects of investigational drugs on those. As a result, novel approaches to tumor size modeling are being developed and applied to support both drug development and regulatory decision making.

  • Characterize safety profile – establish relationship exposure-safety

Optimize dose regimen from a safety perspective and help establishing the therapeutic window.

As for all drugs, the therapeutic window is determined by the balance between efficacy and safety. Exposure-response evaluations of safety data (adverse events, of specific safety findings such as neutropenia) are a crucial element in the submission package to complement the analyses performed on efficacy.

  • Optimize trial designs

Determine dose regimen selection, patient selection, and assessment scheme.

While most oncology modeling focuses on characterizing drug responses to support development decisions, their potential for optimizing clinical trial designs is still undervalued. Apart from supporting selecting dose regimens, models can also assess efficacy and optimize PK/PD measurements to minimize the burden on patients as much as possible.

  • Understand the competitive landscape

Using model-based meta-analysis of (publicly available) clinical trial data for competitors in indication.

Especially in immuno-oncology, compounds are tested across a broad range of tumor types. Model-based meta-analyses enable up-to-date quantification of the competitive landscape in different indications and can help bridge efficacy and/or safety information across indications to support dose setting in a new indication prior to actual clinical testing.


[1] Value of Personalized Medicine, PhRMA report, Spring, 2015

[2] (ASCO website)

[3] “Developing Exposure/Response Models for Anticancer Drug Treatment: Special Considerations,” Mould, DR et al, CPT Pharmacometrics Syst Pharmacol. 2015 Jan; 4(1): e00016.

[4] “Optimizing Oncology Therapeutics Through Quantitative Translational and Clinical Pharmacology: Challenges and Opportunities,” Venkatakrishnan, K, CPT Pharmacology and Therapeutics, Volume 97, No. 1, January, 2015

[5] “Doctors Without Borders,” by Siddhartha Mukherjee, The New York Times Magazine, May 15, 2016

[6] “Quantitative Systems Pharmacology can reduce attrition and improve productivity in pharmaceutical research and development,” by Leil, T and Bertz, R. Frontiers in Pharmacology, Nov. 2014, volume 5

To learn more about M&S in oncology, specifically in immune-oncology, please watch this recent webinar.