The Billion Dollar Molecule, Barry Werth, 1994 - This book is a groundbreaking work in covering a medical science startup since the author was pretty much a fly on the wall during the early stages of starting up of Vertex and had great access on a day to day basis to people within the company. The insights and perspective he builds are hence very compelling. Vertex itself is one of a kind company as its approach to drug discovery is very different from what existing during its early years in the 90s.
My notes -
-
Built to function more like a university style lab than a commercial drug business
-
Everyone understood it would take at least 25 years to taste success. Merck wasn’t built in a day. It was a research powerhouse in 40s and 50s but tasted success only in late 20th century with a string of first-in-class and best-in-class drugs
-
Vertex raised $10m with no products, no revenues and burnt $75k a week. It would take a dozen more years and at least $250m to develop its first drug
-
First project would be to improve upon experimental drug FK-506. It suppressed the immune system to facilitate organ transplants but was extremely toxic. Vertex wanted to go about it in a different way. They wanted to design the molecule atom by atom - not by sifting and sieving through hundreds of thousands of molecules through trial-and-error
-
Academic and industrial scientists tended to be dismissive of each other. Former thrives of publication and publicity while latter has to keep their best works secret.
-
Vertex had several sub-disciplines - of molecular biology (deals with function), chemistry (structure and mechanics), medicinal chemistry, x-ray crystallography, nuclear magnetic resonance spectrography, molecular modelling, computational chemistry, protein engineering, protein chemistry, enzymology etc. (all united by a culture of enlightened self-interest)
-
Glaxo jumped from 25th to 2nd place in the drug industry post success of anti-ulcer drug Zantac
-
Prior to WW-II only a handful of drugs worked - derived from a combination of luck and empiricism. Post penicillin, lab-coated scientists started cooking obscure dirt samples in fermentation broths and screening them for activity.
-
There’s no objective reason why a fungus a million years older than us should be able to lower cholesterol in us and yet it did - as with Merck’s Mevacor ($1.6b drug)
-
By building molecules one wants, rather than fishing for them in nature, we could minimise the side-effects of molecules found in nature
-
Immunosuppression occurs when some but not all of body’s defences are disarmed (thus helping in body accept the transplant and reject the graft). By modifying Fujisawa’s FK-506 from scratch, they could have proof-of-concept that could then be used to build molecules for Multiple sclerosis, juvenile diabetes, rheumatoid arthritis, crohn’s, psoriasis, lupus and other autoimmune diseases
-
Vertex’s idea was to make keys that fit the locks - the locks are always proteins and Vertex had great knowledge of FK-506’s protein receptors than anyone else. Using X-ray crystallography and precisely mapping the floppy protein with thousands of atoms, they had to design a drug that could fit into the receptor at a 1 in 10 billionth of a meter precision (FK-506 protein is FKBP or FK-506 binding protein discovered at Harvard. FK-506 itself was a drug occurring in nature discovered by Fujisawa)
-
To build a molecule that mimics FK-506 built from the ground up, FKBP was essential and could only be sourced from the spleen of human cadavers
-
Molecule on the surface of T-cells on helped the body distinguish self and non-self and trigger the immune reaction (so for transplanters., they had to figure out how to disable them so the body doesn’t reject the organ). Sometimes patients had to take immunosuppressants for life as the threat of rejection persists for life. Cyclosporine was the first and FK-506 the second that helped with this (both severely toxic). Cost-wise former was $240k and latter cheaper at $130k (establishes economics clearly)
-
Most drug companies threw legions of chemists at a problem to change an atom here, a sub-group there in a natural compound and hope it gives better results. Merck was routinely able to bludgeon competition by throwing more chemists at a problem (Vertex despised this approach)
-
Structure based drug design (SBDD) was to design small molecules with only the area of activity (and not the whole large naturally occurring molecule) - with the minimal size and weight possible from the ground up (Orforglipron which is a small molecule mimicking GLP-1R also likely works this way is my guess)
-
Diseases are not treated first and foremost as diseases but as markets
-
A German dye chemist made a dye from coal tar (sulfanilamide) which led to sulfa drugs to combat bacterial infections (1935)
-
Though penicillin was discovered in the 30s, it couldn’t be mass produced until much later as extracted, purification were difficult and yields were poor. Similarly streptomycin was discovered from soil while actively screening and looking for it (unlike penicillin) and could combat TB (1943)
-
Drug companies screening for samples all over the world and even paid half airfare (Squibb) for their employees to bring back samples from where they were vacationing. Cephalosporin was discovered in sewage sludge!
-
Tishler working for Merck synthetically manufactured Cortisone in a 37 step process (1944) and shot Merck’s success into the stratosphere as it was used for all sorts of miraculous ends - from asthma to ulcerative collitis, almost 28 illnesses were treated with it
-
Merck began screening 50k microbes a yea at a plant in Spain post success of Cortisone
-
Merck’s success was on information - not intuition or insight (hence Boger left Merck to start Vertex)
-
What would cost Fujisawa $1000/gm would cost Vertex $1000 a tankload if successful (small molecule at scale)
-
Molecules that look at act like peptides but are structurally different - peptide memetics (Orforglipron again I think is one such)
-
After Linus Pauling’s discovery of the way proteins folded, synthetic chemistry fell in stature to structural chemistry because it could explain form and function instead of blind synthesis
-
Despite Merck’s ability to make Cortisone, Syntex, a little known biotech figured out it could make it from Mexican yam and went on to become largest maker of Cortisone in the world
-
Periplanone-B a pheromone that attracted cockroaches was synthesised and could deceive cockroaches into a frenzy (complex 10 member ring compound)
-
New drugs are rare, novel ones rarer. Most of variants and work on known receptors. Very few novel ones work on a different pathway
-
Medicinal chemists vs process chemists - former discover drugs while latter improve ways of cheaply producing them. Process chemists struggle for respect in companies like Merck
-
Scientists publish to establish new work and authorship. Within industry they do also do this to frustrate competition and entice investors
-
Altruism has no other evolutionarily justified motive but self-interest
-
When researchers found good rewards for doing science that was visible, the attention brought money and money, more research (cancer research was hot on Wall st. in the 90s)
-
A prophet who turns out right is a seer, and wrong, a charlatan
-
Avg. third year revenue for a new drug is $300m
The drug in question eventually did not see the market but Vertex went on to make few other very successful drugs and recently even got approval for a non-opioid painkiller (incidentally, I am reading “Empire of Pain” on the Oxycontin saga which was a Opioid painkiller). This is a field I knew very little about but the more books I have read, the better the clarity is in my head and simpler the mental models to use as an investor in this space. However, this is not a book for everyone. It is big, wordy and dry and not something most people will enjoy. 8/10