To attempt to answer Lisa Marasco's penetrating questions as to how
drugs actually enter milk is rather difficult,  since it is largely an
enigma.  Certainly,  some drugs, particularly lipid soluble drugs probably
enter milk at higher levels during synthesis of milk, and some probably by
dissolving themselves with the actual lipid droplet.   Other more water
soluble drugs probably just traverse the secretory epithelial cell
membranes and enter milk that way.  As of yet, no one has adequately
documented the relative rate that drugs enter milk as a function of milk
synthesis.  But I generally believe that milk acts as a compartment, just
as any other compartment,  and remains in equilibrium to some degree with
the plasma which feeds this wondrous system.  It has been shown that as
plasma levels of drugs rise, so does the level of said drug in milk.  Also,
it has been shown that as the plasma levels drop,  the amount in milk also
drops.  One study with ethanol,  definitely indicated a bidirectional flow
of ethanol,  even without emptying the breast.  I have often thought it
would be wonderful to have an equation, with which one could enter various
numbers such as volume of distribution, half-life, etc.  and come up with a
stated quantity of drug/ml of milk.  Unfortunately we don't have that,
there is just too much variability as of yet.
        I am particularly enamored with protein binding, since it seems to
provide the best data yet as to the relative ability of a drug to penetrate
milk.  Milk:plasma ratio is rather poor,  since it is solely dependent on
the plasma level of the drug. Look at sumatriptan,  which has a M:P ratio
of 7 to 8,  but only nanogram levels actually enter milk.

        It is still quite accurate to recommend that lactating mothers take
medications only after breastfeeding,  so that a longer feeding interval
occurs in an attempt to expose the infant to reduced maternal plasma
levels, and hence, reduced milk levels.  Unfortunately,  there are many
unanswered questions in the physiology of milk.  Because breast milk
production is such as dynamic process, changing from minute to minute,  I
believe it is unlikely we'll ever have mechanisms for forecasting drug
concentrions in milk,  other than prior studies done in humans.

TWH