Chemistry Reference
In-Depth Information
discovery-to-development transition became bidirectional: molecules moved forward
into development, as before
first, the expertise gained further downstream was fed
back into discovery groups and brought to bear on go/no-go decisions at (or shortly
before) the candidate selection stage.
Thanks to the Development 2.0 approach, the transition from discovery to devel-
opment is much less fraught than it used to be. The idea that these two stages require
some collaborative, bidirectional information
—
but
flow to succeed has gained widespread
acceptance [2]. More recently, a number of
firms have begun to scale up these studies
with high-throughput, miniaturized, or automated physicochemical screening during the
hit-to-lead and lead optimization phases of discovery [3
8]. Toxicity, solubility,
pharmacokinetics, ease of manufacture, and drug disposition are among the areas studied
in these
-
programs. The data collected have been assembled into databases
that, in turn, guided the development of several
a posteriori
empirical rules of thumb, like
Lipinski
“
multivariate
”
cation System [10].
Strategically speaking, in each case, access to development stage expertise gives
discovery teams and managers advanced notice of problems and opportunities down-
stream. With their improved awareness, they can make better informed decisions about
which compounds to move into development.
Still, the later stages of preclinical and clinical development can pose further
problems that early development expertise may not recognize. For example, bio-
pharmaceutical properties in humans are notoriously dif
'
s famous
“
Rule of 5
”
[9] and the Biopharmaceutical Classi
cult for even specialists to
predict before they are observed in clinical trials. Likewise, scale-up at the manufacturing
plant (even at clinical scale, never mind commercial) often proves much more complex
than just running the same reactions in a bigger kettle. Crystallizations, in particular,
often become problematic at scale. Finally, while pharmaceutical
firms under Develop-
ment 1.0 once spent tremendous effort trying, in vain, to develop undevelopable mole-
cules, the opposite problem may apply in Development 2.0. In other words, molecules
that were otherwise promising
—
that could, for instance, have high binding af
nity and
speci
have been written off because of their poor stability, or
biopharmaceutical properties, or scale-up potential. Some of these failures may have
been premature. However, even in a Development 2.0 model, expertise in materials
science, biopharmaceutics, and manufacturing is rarely made available to decision
makers early in R&D.
This shortcoming led some of those in preclinical development to a conclusion:
pharmaceutical R&D organizations stand to benefit from taking the salutary develop-
ments of Development 2.0 to their next logical level. In other words, the binary
relationship of preclinical and early development groups should be expanded to
include expertise in
every
phase of drug development. It should also be applied
earlier in discovery research, including the lead optimization phase when molecules
are crafted and selected for further testing. Decision makers should be able to
recognize not only whether a molecule can be readily rendered as a drug product,
but also whether an alternate material form could be a path forward for an otherwise
undevelopable scaffold. These same decision makers should be able to estimate
biopharmaceutical properties and be prepared in advance for any dif
city for their target
—
culties in
