QSP: Strengthening First in Human Dose Selection for Immunotherapy

Selecting the optimal first-in-human (FIH) dose is one of the most challenging steps in drug development. Traditionally, toxicologists and clinical pharmacologists have relied on two metrics to determine the maximum recommended starting dose (MRSD): NOAEL (no observed adverse effect level) and MABEL (minimal anticipated biological effect level). While these measures provided a starting point, experience and evolving science have exposed their limitations, especially in the era of molecular-targeting immunotherapies. Today, quantitative systems pharmacology (QSP) offers a more comprehensive approach, and partnering with experts like Simulations Plus can enhance both safety and efficacy in dose selection.

Traditional Dosing Approaches

For years, MABEL and NOAEL have been the backbone of determining MRSD for FIH study designs. MABEL estimates the minimum dose that induces a measurable biological effect, while NOAEL identifies the highest dose with no observed adverse effects in preclinical studies. These metrics work reasonably well in the context of chemotherapy which has narrow therapeutic windows, where predictable dose-response relationships and established toxicity profiles make translating animal data to humans relatively straightforward.

However, the situation changes dramatically with targeted immunotherapies. Unlike traditional chemotherapeutics, many immunomodulatory agents interact with complex immune networks. Small dose variations can trigger significant, nonlinear responses, widening the therapeutic index (Figure 1). For instance, the same immunotherapy that drives an effective anti-tumor response might also induce severe adverse events, such as cytokine release syndrome. Relying solely on MABEL or NOAEL, which are empirical and based on animal data, may therefore fail to predict critical nuances of human responses.

Inadequacies of MABEL and NOAEL in Targeted Immunotherapy

There are three major areas of inadequacy that MABEL and NOAEL exhibit in targeted immunotherapy.

Overreliance on preclinical data: MABEL is often derived from in vitro potency assessments and scaled animal data, while NOAEL is established from non-human primate (NHP) studies. Yet, differences in physiology, metabolism, and antigen can limit the accuracy of these metrics when applied to humans, resulting in MABEL underestimating (see Figure 2) and NOAEL overestimating the MRSD. A notable example of the latter was the 2006 incident with the CD28 super-agonist antibody TGN1412, where six human volunteers became critically ill-a failure not predicted by animal models because of limited cross-species antibody reactivity. This event underscored the need for alternative methodologies such as QSP.

Static endpoints in a dynamic system: Both MABEL and NOAEL reduce complex pharmacological behavior to a single value-one indicating a minimum effective concentration and the other a toxicity threshold. They do not capture the evolving interplay between a drug's pharmacokinetics (PK) and pharmacodynamics (PD) over time, which is essential for understanding drug behavior in the human body. In immunotherapy, factors like receptor saturation, feedback mechanisms, and cross-talk among cellular pathways require a dynamic approach to dosing.

Lack of mechanistic insight: By ignoring the underlying biological networks, MABEL and NOAEL provide limited information about the drug's mechanistic action. This oversimplification can lead to underestimating off-target effects or delayed toxicities and potentially enforces a one-size-fits-all dosing strategy that might not account for individual patient variability.

The QSP Advantage

QSP represents a paradigm shift in FIH dose selection. By harnessing mathematical models and computational simulations, QSP integrates PK, PD, and detailed biological mechanisms to offer a dynamic and realistic picture of drug behavior. QSP provides a better estimate of:

• Mechanistic integration: Instead of relying on static values from preclinical experiments, QSP incorporates the underlying biological processes that govern drug responses. For example, QSP models can simulate how an immunotherapy interacts with various immune cells over time, capturing both immediate and delayed responses.
• Dynamic simulation: Traditional approaches offer limited snapshots of drug behavior. In contrast, QSP generates dynamic simulations, forecasting both therapeutic outcomes and potential adverse events. This enables clinicians to predict delayed toxicities or receptor saturation effects and adjust dosing strategies accordingly.
• Bridging the preclinical-clinical gap: Translating animal study outcomes to human trials is particularly challenging for immunotherapies. QSP models integrate data from both preclinical studies and emerging clinical data, allowing for continuous refinement of predictions as more human data become available.
• Personalization: Immunotherapies vary greatly among patients due to differences in genetics, the tumor microenvironment, and individual immune status. QSP's ability to incorporate heterogeneous data allows for personalized dosing strategies that are far superior to the blanket approaches offered by MABEL or NOAEL.

Comparing Immunotherapy and Chemotherapy

Chemotherapy typically follows a predictable dose-toxicity curve, where maximum tolerated doses and off-target effects are well characterized. In contrast, immunotherapies modulate complex immune responses that are less predictable. Consequently, what might be an effective dose for one patient may lead to toxicity in another. This variability highlights the limitations of one-dimensional approaches like MABEL and NOAEL, and underscores the need for sophisticated, dynamic modeling via QSP.

The Role of Simulations Plus

For companies seeking to adopt QSP, partnering with expert scientists offers significant benefits:

• Expertise: With decades of experience in pharmacometrics and systems biology, the QSP scientists at Simulations Plus are seasoned experts capable of building robust models that accurately capture the multifaceted behavior of therapeutic compounds.
• Custom frameworks: Recognizing that each drug development program presents unique challenges, we provide customizable QSP solutions that are transparent, reproducible, and aligned with regulatory expectations.
• Accelerated decision making: By simulating multiple dosing scenarios ahead of clinical trials, QSP can quickly identify optimal dosing regimens, reducing the risk of adverse events and expediting the overall drug development timeline.
• Regulatory alignment: Agencies such as the FDA and EMA increasingly value model-informed drug development. The QSP frameworks we develop help justify dose rationale during regulatory submissions and de-risk development initiatives.

While MABEL and NOAEL have served as traditional tools for FIH dose determination, their limitations are increasingly evident in the context of immunotherapy. The dynamic and multifactorial nature of immune responses requires a modeling approach that goes beyond a simple threshold value. QSP, with its integration of mechanistic insights and dynamic simulations, not only bridges the translational gap but also offers the flexibility needed for personalized therapy.

For drug developers, embracing QSP means moving away from the one-size-fits-all mindset, toward a more nuanced understanding of dosing that can enhance both patient safety and therapeutic efficacy. As immunotherapies continue to evolve, partnering with QSP experts becomes essential for unlocking the full potential of innovative treatments. This shift towards QSP promises to pave the way for next-generation clinical trials-safer, faster, and more attuned to the complexities of human biology.

By integrating advanced simulation tools and embracing model-informed strategies, the future of FIH dosing is not only more predictable but also tailored to the individual needs of patients. It's time to leave behind the limitations of MABEL and NOAEL and adopt a forward-thinking approach that leverages the power of QSP.

If you'd like to learn how your team can begin using QSP for immunotherapy research, let's talk.

Discuss My Project

Relevant Sources and Further Reading:

1. Hojouj M, Landers D, Cruz R, Stuart M (2024) Project Optimus: An Overview of the Principles and Challenges. Journal of Immunology Research & Reports. SRC/JIRR-147. DOI: doi.org/10.47363/JIRR/2024(4)136
2. https://www.simulations-plus.com/resource/phasing-out-animal-testing-responding-to-fda-and-emas-strategic-shifts/
3. Muller PY, Milton M, Lloyd P, Sims J, Brennan FR. The minimum anticipated biological effect level (MABEL) for selection of first human dose in clinical trials with monoclonal antibodies. Curr Opin Biotechnol. 2009 Dec;20(6):722-9. doi: 10.1016/j.copbio.2009.10.013. Epub 2009 Nov 5. PMID: 19896825.
4. Attarwala H. TGN1412: From Discovery to Disaster. J Young Pharm. 2010 Jul;2(3):332-6. doi: 10.4103/0975-1483.66810. PMID: 21042496; PMCID: PMC2964774.