Until the early 1960s, lead compound optimization was an intuitive endeavor based
on long experience, keen observation, serendipity, sheer luck, and a lot of hard work.
The probabilities of finding a clinically useful drug were not good; it was estimated
that anywhere from 3000 to 5000 compounds were synthesized in order to produce
one optimized drug. With today’s even stricter drug safety regulations, the proportions
are even worse and the costs skyrocket, retarding the introduction of new drugs to an
almost dangerous extent. The classical method usually applied in lead compound opti-
mization was molecular modification — the design of analogs of a proven active
“lead” compound. The guiding principle was the paradigm that minor changes in a
molecular structure lead to minor, quantitative alterations in its biological effects.
Although this may be true in closely related series, it depends on the definition of
“minor” changes. The addition of two seemingly insignificant hydrogen atoms to the
∆^8 double bond of ergot alkaloids eliminates their uterotonic activity, but replace-
ment of the N-CH 3 substituent by the large phenethyl group in morphine increases the
activity less than tenfold. Extension of the side chain of diethazine by only one carbon
atom led to the serendipitous discovery of chlorpromazine and the field of modern
psychopharmacology.
There are two conclusions to be drawn from these random examples. First, a merely
structural change in an organic molecule is meaningless as long as its physicochemical
consequences remain unexplored and the molecular basis of its action remains
unknown. Structure, in the organic chemical sense, is only a repository, a carrier of
numerous parameters of vital importance of drug activity, as is amply illustrated in the
first chapter of this book.
The second conclusion to be drawn from the above examples—and innumerable
others—is that the discovery of qualitatively new pharmacological effects is often a
discontinuous jump in an otherwise monotonous series of drug analogs and is hard to
predict, even with fairly sophisticated methods.
Although a beginning has been made, drug design is far from being either automatic
or foolproof. The decision to optimize a proper lead compound—a necessity in drug
design and development—is still based on experience, serendipity, and luck, given our
basic ignorance of molecular phenomena at the cellular level. Now, however, in the 21st
century we can at least have some confidence (thanks in no small part to computer-
aided molecular design and bioinformatics) that the optimization of lead compounds
and the corresponding discovery of new drugs will be able to keep pace with the
progress of biomedical research.
3.4 OPTIMIZING THE LEAD COMPOUND:
THE PHARMACODYNAMIC PHASE
Once a lead compound has been identified by one of the techniques described above
and the decision has been made to optimize this compound, the next task is to deter-
mine an approach to compound optimization. Typically, this approach is achieved via a
two-step strategy:
- Optimizing the compound for pharmacodynamic interactions
- Optimizing the compound for pharmacokinetic and pharmaceutical interactions
134 MEDICINAL CHEMISTRY