I want to leave the board with this explanation from our lead formulator and scientific consultant.
I can use this as an example. Ican point out plenty of molecules which make great substrates for COMT. I cango even further and point out that they were then confirmed in vivo to besubstrates of COMT (which, by the way, is not the only way these compounds aremetabolized as is nearly always the case) based upon urinary metabolites. Yet,simply being a good substrate for COMT does not equal 0% bioavailability. Itdoes not equal a 1 minute half-life. It does not mean, in any way, that it willhave the same pharmacokinetic properties as another compound, even if thestructural similarities are many; in fact stereoisomers (thus, structuralformula is not just similar but actually identical) of various compounds, includingsome beta 2-adrenoceptor agonists themselves, display substantial differencesin bioavailability and other pk parameters. This is why these parameters mustbe determined experimentally, period. This is why drug discovery departmentsdon't have a guy on a computer in the corner, google searching for similarlooking molecules, reviewing their pharmacokinetic properties and thenconcluding that the compound in question is probably the same, so they ditchit. You can form a hypothesis that they may be similar in that regard, but toknow for certain, you must investigate through experimentation. That is basicscience. Anything else is a prediction and will always remain so, period. Andany person that doesn't have an advanced scientific background that mightbelieve that such comparisons just make good enough sense that theirhypothesis, "just HAS to be right", should visit one of the severaljournal publications (i.e., negative result journals) that focus on nothing buthypotheses that were not supported by their experimental results. Of course, Ishould also point, most don't bother to ever publish such results, so let'sjust say the vast majority of hypotheses don't fit with the experimental dataand that's because we're human and we're studying what are arguably complexmatters where a multitude of variables aren't known and those that are may notbe taken into account.
In this particular case, simply being a substrate for a given enzyme whichserves to create inactive metabolites does not mean A) it can't be saturated B)that the amount produced is enough to yield none of the active base C) that itis the major route or the only route of metabolism and D) that it isn'tresistant to another route E) that all routes of metabolism produce onlyinactive metabolites as opposed to equipotent or even hyperpotent metabolitesF) It really does go on and on regarding potential variables at work here whichis why experimentation is vital.
In fact, we have a compound which we are investigating for potential futurerelease which has the catechol ring (thus it was supposed that it would likelybe a substrate for COMT) and yes, it does in fact undergo o-methylation to asubstantial degree, as it has been confirmed by determination of theo-methylated urinary (there's that experimentation part again). So, using thispredictive logic, this compound should be thrown in the trash despitedmetabolites emonstrating potent activity at the beta 2-adrenoceptor. Yet, whenactually investigating the oral bioavailability of this compound(experimentation), we find it is in the double digits. And we find thehalf-life isn't just 1 or 2 minutes but much, much longer. So, we have acompound that is a great substrate for COMT, yet still has pk characteristicsthat make it viable; this is why experimental data are important andsupposition is meaningless without it. Are these pk values ideal? Compared tomany pharmaceuticals, no (although, there are even some examples ofpharmaceutical compounds with less than ideal bioavailability values). But, wearen't a pharmaceutical company and this isn't the pharmaceutical industry. Ouronly similarity is with those pharmaceutical companies that still haveoperating sections of natural products drug discovery, only we generally haveto stop at the lead in most cases. But, often times, depending upon what thegoal is, the lead may in fact be good enough. I would also hope that mostunderstand that even the slightest change in a molecule can yield substantiallydivergent pharmacokinetic and pharmacodynamic properties.
In any case, this brings us to the next point which is that there is more toconsider aside from bioavailability and half-life. If a particular compoundpossesses the necessary potency at whatever your given molecular target may be,the bioavailability in particular may not be of much importance, especially ifthere is room for a variable dose. For example, if the concentration needed foractivity is in the picomolar or nanomolar range and the given compound yieldsserum concentrations in that range after administration of 1,000 mg or 3,000mg, what does it matter if the bioavailability is only 5%? The answer is thatit doesn't.
I think discussion is great and reviews like this are a great way to sparkexperimentation and improve study design, but let's be better than simplytaking a hypothesis and using it as a justification to say whether somethinghas value or not. In fact, if we want to respect science at all, let's not takehypotheses and use them as, "facts", which in reality, do not existto begin with.