Dose Optimization: Getting it Right for Vulnerable Populations

Clinical Researcher—April 2023 (Volume 37, Issue 2)

SCIENCE & SOCIETY

Justin Hay; Frank Engler

Pandemics amplify existing strengths and vulnerabilities—not only highlighting who are among the at-risk populations in our society, but also bringing to prominence strengths and weaknesses within healthcare systems and drug development processes. Advancements in vaccines and antivirals were carried out at unprecedented speed during the COVID-19 pandemic using exciting, new technologies, and there are now a slew of enquiries, reports, and (hopefully) “lessons learned” from those experiences. But is dose optimization one of these teachings?

Hindsight is Perfect

In theory, dose optimization is a relatively simple concept—make sure a drug’s dose delivers the best balance between risk and benefit for the patient—yet in practice it is a different story. Dose optimization has never been perfectly spot on, with dose changes to labels occurring in one of about every five small molecules and in about one of 10 for biologics.{1,2} For small molecules, most label changes are related to safety and due to the dose being too high. However, safety does not appear to be the driving issue for biologicals, where label changes are mostly due to patient convenience and related to efficacy.{1,2}

Understanding the drug’s dose-exposure-response relationship is key to justifying the dose on the label.{3,4} For monoclonal antibodies targeting COVID-19, this is especially important.

Opportunity Knocks But Once

As with other therapeutic fields, such as oncology{5}, there is an absence of dose-response data available for pandemic medicines, albeit for different reasons. Pandemics demand that decisions are made quickly and require medicine development programs and approvals to be ruthlessly efficient because, initially, there is an imminent, unmet need for treatments and later they need to keep pace with the evolution of the viral target. While the doses selected may be justified, are there opportunities to ensure they are optimized even with only limited data available?

Double-Edged Sword

Modelling and simulation have played critical roles in our understanding of the pharmacology of COVID-19 therapies. Their applications range from describing the population pharmacokinetics (PK), to investigating possible dosing regimens and exposures in pediatric populations, to highlighting potential drug-drug interactions using physiologically based pharmacokinetic (PBPK) modelling. Modelling and simulation have also been crucial to dose justification for monoclonals, helping to determine whether the level of antibody is reaching sufficient levels quickly enough in the target tissue and for long enough to offer protection.

However, the challenge for COVID-19 therapies (and indeed many other anti-infectives) is that variants cause big headaches for research and development, especially for the monoclonals targeting an ever-changing spike protein or non-conserved surface antigen. In vitro neutralization curves have been critical to our understanding of the susceptibility of variants to monoclonals, with modelling and simulation providing a critical link between the IC50, the IC80/90 (should it be calculated or measured?), the plasma-to-target-tissue ratio (should it target the lower or upper respiratory tract?), the target threshold, the variability in the PK, and ultimately understanding whether there is enough drug at the site of action.

A handful of monoclonal antibodies have received authorization in the European Union (EU) for the treatment and/or prevention of COVID-19—among them, regdanvimab (Regkirona), casirivimab/imdevimab (Ronapreve, known as REGEN-COV in the U.S.), sotrovimab (Xevudy) and tixagevimab/cilgavimab (Evusheld). Others have received opinions under Art.5(3) in the EU or Emergency Use Authorization in the U.S.

Interestingly for Evusheld, at the time of initial approval, there were limited clinical data supporting the use of tixagevimab and cilgavimab against the newly circulating omicron variant.{6} Regulatory agencies advised that a 600 mg dose, rather than the initially studied 300 mg dose, was needed for pre-exposure prophylaxis of COVID-19 based on the levels needed to neutralize the omicron variant in vitro. Insensitive variants have already knocked casirivimab/imdevimab, sotrovimab, and bamlanivimab/etesevimab out of the race.{7,8} This story is bound to be repeated in the future with a plethora of emerging variants on the horizon.

Meanwhile, there are many new monoclonal antibodies against COVID-19 in development, but making sure that medicines are safe and effective is a time-consuming business. Going forward, is there enough time to wait for a confirmatory clinical trial to tell us whether a monoclonal against COVID-19 (still) works? Or is it enough to rely on some clinical data—modelling the in vitro data, combining them with assumptions of human physiology, and simulating target concentrations in target populations? Is this pushing the limits with regard to bridging nonclinical and clinical development? Regardless of what the answers are, it is important to make sure that any underlying assumptions used for modelling and simulation are robust and accurate.

Recently the U.S. Food and Drug Administration and European Medicines Agency (EMA) hosted a joint workshop{9} to discuss this exact topic with some interesting take-home messages regarding the use of surrogates of clinical efficacy like viral load, neutralization titers, and PK/pharmacodynamic modelling and whether these can be used to predict or support clinical efficacy. It is hoped these discussions can be leveraged to bring forward monoclonal therapies in a rapidly changing world. Not only will this help us with the further development of therapies for COVID-19, it also will hold us in good stead so we can be prepared for the next pandemic and meet any unmet need as quickly as possible.

Eye On the Target

Even though we may have put COVID-19 somewhat behind us, and it does not make the headlines as much anymore, there are many lessons to be learnt or perhaps re-learnt from it. While for many, COVID-19 may result in a mild sniffle and a cough, for the immunocompromised for whom vaccination may not lead to an adequate immune response, COVID-19 still lingers as a significant threat to their health, and it is for them that we need to make sure that we get the dose right.

References

  1. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6422665/
  2. https://pubmed.ncbi.nlm.nih.gov/12426927/
  3. https://pubmed.ncbi.nlm.nih.gov/25773468/
  4. https://pubmed.ncbi.nlm.nih.gov/33049309/
  5. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8586755/
  6. https://www.ema.europa.eu/en/medicines/human/EPAR/evusheld
  7. https://www.who.int/publications/i/item/WHO-2019-nCoV-therapeutics-2023.1
  8. https://www.fda.gov/news-events/press-announcements/coronavirus-covid-19-update-fda-limits-use-certain-monoclonal-antibodies-treat-covid-19-due-omicron
  9. https://www.ema.europa.eu/en/events/joint-ema-fda-workshop-efficacy-monoclonal-antibodies-context-rapidly-evolving-sars-cov-2-variants#registration-section

Justin Hay is a Senior Director at Certara with more than 20 years of clinical pharmacology experience. Career highlights include working as Senior Pharmacokinetics Assessor and Deputy Unit Manager at the Medicines and Healthcare products Regulatory Agency (MHRA) in the United Kingdom. He has also been a member of the EMA’s Modelling and Simulation Working Party.

Frank Engler is a Senior Director at Certara with eight years of experience in clinical pharmacology, focusing on model-informed drug development in the biologics space. He continues to work on several biologic therapies for the treatment of COVID-19.