British Association For Crystal Growth Annual Conference 2017
From Molecules to Medicines the Solid-State Way
S. Reutzel-Edens
Eli Lilly and Company, Indianapolis, IN USA
[email protected]
What would it take to transform a molecule to a medicine maximally, if not purely, in silico? This question is
increasingly being raised as the pharmaceutical industry seeks to deliver innovative medicines to patients
faster, while maintaining the highest standards of patient safety, integrity and data quality. The challenge is
nothing short of designing the right drug product and the right processes to deliver optimal product
performance the first time, and the answer will require both ground-breaking science and disruptive digital
technologies to understand and predict molecular, solid-state, process, product and performance attributes
that lead to desired patient outcomes. Moving to a digital drug product design paradigm is expected to
fundamentally change how drugs are developed as increasing computational power and advanced
algorithms allow more and more meaningful experiments to be run in parallel in a computer. However, the
pharmaceutical industry is far from being able to fully design, much like an airplane or a bridge would be
engineered today, drug products that meet patient requirements without significant trial-and-error
experimentation. Nonetheless, as much progress is being made on several fronts relating to computational
drug substance and drug product design, it is worthwhile to trace the digital path from a molecule to a
medicine (e.g., Fig. 1) to assess the state of the art and see where scientific advances are clearly needed.
Fig. 1: Example of a digital design roadmap to understand, rationalize and predict molecular, solid-state,
process, product and performance attributes that lead to desired patient outcomes.
An essential first step, and oftentimes a bottleneck, in drug product design is the identification of crystalline
forms of the drug molecule, a task which today relies heavily on the effectiveness of experimental salt and
polymorph screening programs. Crystallization provides a means to purify and recover the drug substance
coming out of the final step of the synthesis and to isolate the drug in a crystalline form that is suitable for
downstream processing. It is also used to define the material properties (e.g., stability or solubility) of the
drug substance that will ensure consistency in the safety and efficacy profile of the product throughout its
shelf life. The importance of identifying the thermodynamically stable crystal form early in drug product
development cannot be overstated. In addition to potentially delaying regulatory submission and marketing
approval, a form change prompted by the late discovery of a more stable polymorph (or hydrate) will
inevitably incur significant costs associated with redeveloping new crystallization and formulation processes,
repeating toxicology and stability studies, establishing bioequivalence, and potentially adjusting dose
strengths. The consequences of missing a more stable crystal form only to have it appear after the product is
on the market can be catastrophic, minimally threatening market supply and in the worst case, forcing
product withdrawal (cf. ritonavir).
An ever-expanding range of experimental techniques has been shown capable of producing novel solid
forms, yet it is possible that practically important forms might not be found in the timescales currently allotted
for solid form screening. In this presentation, cases studies are presented highlighting the practical
challenges in defining solid form landscapes and ensuring that relevant forms are not missed. Crystal
British Association For Crystal Growth Annual Conference 2017
structure prediction is explored as a complementary in silico approach to pharmaceutical solid form
screening. This first step toward realizing the vision of digital solid-state form design has helped to establish
molecular-level understanding of the crystallization behavior of APIs, as well as shown the need for further
development [1-2].
References:
[1] S.L. Price, S.M. Reutzel-Edens, Drug Discovery Today, 2016, 21, 912.
[2] S.L. Price, D.E. Braun, S.M. Reutzel-Edens, Chemical Communications, 2016, 52, 7065.