3:15 PM
 | 
Aug 02, 2018
 |  BC Innovations  |  Tools & Techniques

Model firsts

How quantitative systems pharmacology can pick the right dose for first-in-human trials

By combining disparate data into coherent mechanistic models, quantitative systems pharmacology is becoming a key tool for picking the right dose for first-in-human trials and other early make-or-break decisions. Advocates see it as part of an expanding toolbox of models that can yield better safety and efficacy predictions from preclinical data, and want regulators to include it in their guidances.

Moreover, they argue the disastrous outcome of the 2016 Phase I BIA 10-2474 trial could have been avoided if published QSP studies on parallel compounds had been taken into account.

QSP can enable companies to turn data into actionable decisions by translating biological mechanisms and lab measurements into mathematical equations and computational simulations. The models go beyond standard allometric scaling from finite animal data, to incorporate mechanistic information about what cellular compartments a molecule acts in, and how fast it is both produced and degraded, for example.

“The decisions in drug companies we have to make at a high level are -- ‘should we keep this program going forward’ and ‘what are the risks that we’re going to encounter with it?’” said Vikram Sinha, Associate VP of Quantitative Sciences at the Merck Research Laboratories unit of Merck & Co. Inc. “That’s where these models are becoming more and more influential, because this is the only way that we can integrate all this information.”

Certara L.P. believes QSP can provide critical guidance for making decisions on first-in-human dosing, where mistakes can be costly. The company develops software for model-informed drug development and other forms of pharmacology and PK/PD modeling.

In 2014, Certara researchers published a QSP model suggesting the therapeutic efficacy of FAAH inhibitors would reach a maximum at doses of 3 mg. That’s well below the daily 50 mg dose of Bial-Portela & Ca. S.A.’s FAAH inhibitor BIA 10-2474 that induced neurological toxicity in healthy volunteers in the 2016 Phase I trial, leaving one brain dead and hospitalizing five others. Bial-Portela did not respond to requests for an interview.

“Our QSP model predicted that at relatively low dose levels you will saturate the FAAH enzyme with your compounds, but that doesn’t necessarily lead to any increased desired pharmacology,” said Piet van der Graaf, Certara’s VP of QSP.

“QSP actually becomes important in trying to say, how are we going to dose escalate, and how best should we be designing this to actually understand the dose-response relationship?”

Joshua Apgar, Applied BioMath

In a June letter to the editor of Clinical Pharmacology & Therapeutics, Certara argued EMA’s 2017 first-in-human guidelines, published in response to the BIA 10-2474 incident, failed to address the potential of QSP models to make dose selection safer. EMA responded that it “encouraged” and “supported” the use of mechanistic models, while noting it would question models with insufficient validation, particularly those recommending higher doses than standard preclinical approaches. The agency said it would consider including guidelines on QSP in future documents.

Sinha said that while FDA and EMA guidelines have made choosing a starting dose fairly straightforward, QSP models still have a lot to offer as they can address the more complicated question of how far to escalate dosing to capture desired pharmacology. “QSP actually becomes important in trying to say, how are we going to dose escalate, and how best should we be designing this to actually understand the dose-response relationship?”

Applied BioMath LLC CSO Joshua Apgar told BioCentury QSP counters the conventional thinking that “if it’s tolerated, higher must be better.”

“QSP models provide a valuable tool to say, what’s the highest dose that has a rational basis for being maximally efficacious?”...

Read the full 2987 word article

User Sign in

Trial Subscription

Get a 4-week free trial subscription to BioCentury Innovations

Article Purchase

$100 USD
More Info >