How Pharmaceuticals Work

Jul 2013

(Based on a talk at Viralizando)


“If people let the government decide what foods they eat and what medicines they take, their bodies will soon be in a very sorry state.”

– Thomas Jefferson


“Wherever the art of Medicine is loved, there is also a love of Humanity.”

– Hippocrates


It always amazes me how little people know about the pharmaceutical industry. You’d be forgiven for thinking it was just consumers but even people in the industry don’t really know how it works. People like doctors or pharmacists.

Medicine is one of these peculiar industries which affect everyone and has so many colossal moving parts that its hard to grasp.

Most people hate the pharmaceuticals industry but I don’t think they really know why. They hear large amounts of money being spent and they get angry without understanding where the money goes or why.

So first we’re going to break down what exactly the moving parts are. It’s important to understand the scale involved. There are very few industries that are so compliance and regulation intensive, have such large product time horizons, require as much capital or have as large an addressable market than pharma. The market is everyone.

In rough numbers it takes 12-15 years to commercialize a product, each product costs nearly 1.2 billion dollars to create and the creation of products is largely a luck-based adventure into the subatomic workings of people, nature and the universe.


So why does a drug take 12 years to get to market? It’s easier to break it down first into what is a drug? A drug is just a new molecule or compound put into a form that is easy to take. The molecule then interacts with the body in such a way that it fixes your illness. There are two types of drugs; Small Molecule which make up most pharmaceutical products and Large Molecule or Biosimilars which are a new technology.

With Small Molecule, the drug contains microscopic amounts of a synthesized new molecule or compound which then binds to cells near the disease. This then activates or inhibits cell receptors to help the body fight off the disease. Because the compound is synthetic the body rejects the compound and it passes through after curing the disease. This is the standard for medicines.

With Biosimilars, the drug contains living organisms or cells. The living organism then become part of the person who takes the drug. Because the drug contains living matter it is much harder to manufacture and transport but is cheaper overall to make because you don’t have to buy chemicals since you can artificially create and replicate organic matter. A good way to think about how Biosimilars work is by thinking about how vaccines work.

The key part is the new molecule. The molecule first has to be discovered. That is more difficult then it sounds. Discovery is not really the right word because frequently it is at your fingertips the entire time because all molecules are a function of how well we understand the world.

Just 50 years ago it wasn’t well understood that the immune system fights infection and one way to beat diseases is to boost it. So it is like putting the universe under a microscope to better understand how everything works. You first have to know how something is made before you can go about fixing it when it breaks.

New molecules can come from anywhere. One of the active ingredients in Aspirin is Salicylic acid which comes from the bark of Willow Trees. One drug is based on a synthesized molecule from the regenerative properties of the stomach lining of a platypus. Or new compounds which are just a bunch of other compounds and molecules mixed together in different amounts. Most of these new molecules and new compounds come from research and is why funding research is so important.

How many new molecules or compounds are discovered per year? Around 6,000 completely new chemical substances for each new drug. So only 1 in every 6,000 new compounds makes it to market. But that’s because not every new compound has healing properties, only some of them.

This funding frequently comes under criticism since there is no good heuristic for their results like the number of discoveries versus the amount of money that goes into R&D. It’s not like you put X amount of money in and you get 3X worth of discoveries. Some very significant discoveries come from small labs and a lot of very well funded institutions don’t find anything at all.

This is because research is sometimes called just a series of very fortunate accidents. It might take a scientist a decade to have the knowledge and expertise to recognise an accident as a genuine discovery. Or it might take decades of work to create the circumstances for a discovery to be made.

Very famous examples include Viagra which was originally intended to reduce blood pressure before other uses for it were observed. Pennicilin came from biologist Alexander Fleming watching a sandwhich become mouldy after he accidentally left it overnight in a laboratory. A biologist drinking red wine in a lab accidentally spills it into a petri dish leading to the discovery of Resveratrol. Or Jonas Salk who worked on a cure for polio for nearly a decade before discovering the Salk Polio Vaccine.

Now the more money that goes into research, the more frequent these discoveries happen. Roughly 5 billion dollars is spent on research for every new drug successfully commercialized and this level of sustained capital is required to create more cures.


So a good way to think about the lifecycle of the industry is a molecule is first discovered by a scientist. The scientist might work at a university or a private laboratory funding the research. That molecule is then extracted and synthesized for efficacy at treating a disease in a lab. A bit like a chef mixing ingredients until the right recipe is made.

Once the recipe is right, a New Molecular Entity or NME and Investigational New Drug Application or IND is filed with the Food and Drug Administration or FDA. It’s a bit like saying we think we might have discovered something important but need to test further. The FDA are responsible for the approval of new health products onto the market. Their job is to make sure unstable compounds and potentially harmful or unsafe drugs don’t reach the consumer.

Every part of the world has their own version of the FDA. In Australia, it’s the Therapeutic Goods Administration or TGA. In Europe it is the Europeans Medicines Agency or EMA. And a lot of these have cross approval. If something is approved by the FDA, it’ll be approved faster by the TGA and EMA.

Just to file an IND is over a million dollars which is why small pharmaceuticals aren’t able to create new drugs effectively. If the IND is approved, the new drug goes into clinical trials. There are 4 phases to clinical trials.

Phase 1 or preclinical trials is animal testing such as rats, monkeys, dogs or any animal with similar pharmacokinetics to humans. Pharmacokinetics is just a fancy word which means how they absorb and react to a drug. If there are no negative effects and works as intended it goes to phase 2.

Phase 2 clinical trials is testing in tens or hundreds of healthy volunteers to establish a baseline. And in tens or hundreds of actual patients or the people for whom the drug is designed for to see if it actually works. If there are no negative effects and it passes phase 2, it goes on to phase 3. If a drug is going to fail, it should fail in phase 2 because phase 3 is when it becomes really expensive. 90% of the cost of clinical trials occur in phase 3.

Phase 3 clinical trials has to be done with nearly 2000 patients exhibiting the disease. Except for cancer or heart drugs which have to go through phase 3 studies with nearly 10,000 patients. Finding these patients and administering the drugs is very time-consuming and very expensive, measuring in the hundreds of millions of dollars and can take years to complete.

The results for this is compiled and a New Drug Application or NDA is filed. This costs another few million dollars and is given to the FDA. They then decide whether the drug is safe enough and works well enough for regular consumers. If it is approved, a patent is filed and it starts being manufactured for general use.

A lot of pharmaceuticals are also manufacturers but a lot are just labs and they have to contract the making of drugs to other companies that are manufacturing only. The manufacturing happens in a factory that has to have high standards of quality control and is regulated by the World Health Organisation or WHO.

A good way to think about the manufacturers is kind of like a restaurant. These manufacturers are given a recipe by the drug company then they buy the chemical ingredients from chemical companies like Citigroup or Exxon or DOW Chemicals and then mix it in the right quantities to create the drug.

The drug then gets packaged, branded and shipped around the world. The drug is sold to wholesalers and distributors. Those wholesalers and distributors increase the cost to make a profit then sells them in bulk to pharmacies and retailers.

Each of those pharmacies and retailers increases the cost further to make a profit and sells it to the consumer. Usually this step is when the highest cost increase happens because retailers can charge whatever they want. This is how you can buy a drug for $15 in one place and $115 just down the road.

Every step in this process usually requires some form of new licensing from manufacturing licences to import licences to pharmaceutical licences to pharmacy licences. And the consumer is not allowed to purchase the drug unless a doctor writes a prescription to. They then run post-clinical trials to see the long-term effect of the drug in the public and is sometimes called phase 4.


The failure rate of this sustained capital is very high. Only 20 of every 6,000 new compounds makes it to clinical trials and only 1 in every 20 drugs that make it to clinical trials are eventually approved. In any one year, the FDA approves about 20 new drugs with 39 being approved in 2012. The most approved in a single year since 1996 which had 53 approvals. The lowest was 17 approved in 2006.

So as a conservative rule of thumb, the entire pharmaceutical industry is geared to commercialize only 20 drugs each year. And the results people see from these drugs are about 12 years worth of compounding research, studies and development. That’s why any major changes to the industry, the effect isn’t seen until a decade later and is why changing the price of drugs is actually very difficult to do.

These 20 new drugs have to recoup the entire capital expenditure for the entire industry. This is when the clock starts. Because a company only holds a patent on a drug for 20 years. If they spent 1.2 billion dollars to bring it to market, they have to earn that back before the patent runs out.

But not just the cost of the drug itself but on all R&D costs involved. Remember only 1 in 20 drugs make it through clinical trials? And only 20 of every 6,000 new compounds even gets there in the first place?

That one drug has to recoup the cost of all 19 of the failed ones and all of the research that went into all 6,000 new compounds that never made it to clinical trials. This number is very large and estimated to be around 4 billion dollars. That’s a lot of $19 boxes of Lipitor.

But that’s just to recoup your costs, right? You’re a company and you also have to make a profit if you want to stay in business. That profit has to be very large if you’re going to invest 4 billion dollars just to have the opportunity to make a profit. So when people complain about the high cost of drugs, they don’t see the costs involved to get there.

But that’s just the cost of creating the drug. Next comes the rollout of it. This is also very expensive and is frequently overlooked when people talk about drug commercialization. Nobody actually buys their drugs from a drug company, they buy it from pharmacies. The drug has to be delivered to every pharmacy in the country. If you are approved by the FDA, TGA and EMA which most popular drugs are, that means delivering the drug to every pharmacy in all of Europe, Latin America, Australia and the United States. This is a logistics nightmare.

But a sick person can’t just walk into a pharmacy and buy a drug, they need a prescription first. The only people who can write prescriptions are doctors. If this is a new drug, most doctors don’t know about it, what it does or what it cures – so how could they prescribe it? So the next step is educating doctors. That was the original purpose of medical journals. To keep doctors up to date on the latest drugs, diseases and medical practice.

But there’s a small flaw here. It’s based on the assumption that all doctors are up to date on new technologies. This most certainly is not the case. And that all doctors will immediately switch to prescribing this new drug, bearing in mind they already have a way of treating whatever disease this new drug cures. So you first have to convince the doctor that your drug is a better way of doing it.

Doctors by virtue of their title are usually very smart. Try explaining to a smart person to change what they’ve always done for years. Multiply that by every doctor on 4 continents and you see the extent of the problem. If you don’t get doctors to prescribe the drug, then it won’t make sales. And you have to do all of this soon because your patent is expiring in 20 years. It’s why frequently pharmaceuticals start educating doctors on a new drug while it is still in clinical trials, expecting it to take years before it reaches the market.

It also means certain types of diseases aren’t big enough for a drug to be manufactured to cure them. A lot of diseases actually have cures but the addressable market for them is not large enough to make them commercially tenable. The cost is still the same regardless of whether a drug treats cholesterol or gum disease. But a lot more people have high cholesterol than have gum disease.

So there is a deadpool of diseases that have cures but effect so few people that there is no profitable way to bring their cures to market. Such as types of arthritis or types of HIV. Not enough people suffer from the disease to buy the drug so the drug won’t get made.

Figuring out how to square that circle would be one of the most beneficial things for humanity. Perhaps by incentivising smaller pharmaceutical companies to actively go after niche diseases. To score lots of base hits instead of just going for home runs.


A lot of people wonder why pharmaceutical companies spend so much money on lobbying, marketing and keeping physicians on retainer. These activities are largely why they are considered evil. But if you looked at it as an industry snapshot or proportion of total expenses then it is not really that much anyway.

When people complain that drug companies spend more on marketing than on R&D is kind of a misnomer. This isn’t actually a bad thing because the point of marketing is to bring in more revenue which then funds more R&D. The pie is then larger.

What people don’t realise is this is actually a result of desperation. When the patent expires on one of your flagship products, you can’t just churn out a new product quickly. And when that product represents billions of dollars of revenue, you have to resort to other ways of making money to make up for the lost margins. It’s not out of greed, it’s out of desperation. If you don’t, you’ll go out of business.

If you own something and you know at some point it will run out, then you will try to make the most of it before it does. That’s what is happening. When a patent expires on a drug, the drug company loses it forever so is trying to make as much money as they can before that happens because creating new drugs is hard and expensive.

One of these is to get doctors on payroll to prescribe certain kinds of drugs so you make more sales. Another is to get people to want to buy certain kinds of drugs via direct to consumer advertising to make more sales. The third is to bully the government into letting them charge more for it and change the rules by lobbying.

This is very telling. When a pharmaceutical company realises it can increase its margins more by lobbying than it can by spending money on R&D and creating products is indicative of another type of problem. 1) that the cost of drug commercialization has gotten too high to be viable and 2) that the government is corrupt and easily bullied.

I don’t know how to fix number 2, it urgently needs to be though I would have no idea where to begin. But number 1 seems reasonable. A more important question is how would you reduce the cost of commercializing a drug? This is when you come to an interesting catch 22.

It is largely because of the approval process that makes drugs so expensive to produce and their cost increasing. In 1975, the cost of research and development on a drug was 100 million in todays dollars and medicines were relatively inexpensive. Today it is 1.2 billion dollars and the price of medicine has gone up proportionally.

The very people who approve new drugs are the FDA who are government run and susceptible to lobbying. They’re like the gatekeepers to the entire industry. Some people argue that the process of dealing with an approval body like the FDA is so capital intensive and bureaucracy prone that it’s not worth the effort. The FDA who are the same people responsible for approving new drugs are partially the reason why drugs cost so much.

I wasn’t sure what specifically about the approval process accounts for the 10X increase in costs, so I asked Derek Lowe, the popular chemist and blogger for his thoughts. He said:

“I think it’s a combination of factors. The FDA wants better statistics, so the number of patients for any trial goes up 1X. Then the amount of documentation and data collection goes up 2X. We’re going after harder diseases, so that sends up the number of patients, too, by a factor of 3X. This lengthens the timeline of the trials by 4X. Keep in mind that we’re also trying to do a better job with patient selection, which is supposed to increase the chances of the trial working, but that’s pretty much canceled out by the difficulties of the disease, and that adds a further paperwork/delay expense of 5X. And by the time you multiply all those X factors together, costs have gone up by a really nasty amount.”

If a drug is rejected by the FDA, the company behind it is effectively dead, because they will have spent hundreds of millions of dollars developing a product with no way to bring it to market. In 2010, an anti-obesity drug Lorquess by Arena Pharmaceuticals was rejected at phase 3 clinical trials. That same day their stock price dropped over 80%. This is more common than people think. About 42% of drugs that make it to phase 3 clinical studies fail representing billions of dollars of wasted investment.


Currently there is no way for a pharmaceutical to test a drug with large numbers of people without going into phase 3 clinical studies which requires hundreds of millions of dollars of investment. If after these studies, the drug is rejected then that cost is largely wasted. So the way to fix this is for regulators to reduce the cost of phase 3 clinical trials.

And for more flexibility in the approval process. Most diseases effect the body in similar ways. A lot of drugs that fail phase 2 studies might be effective against other similar diseases but are never trialed. This was the case in the discovery of Viagra which was initially used to treat blood pressure, or Pseudoephedrine which was initially used for sleep related disorders, or Thalidomide the horror drug when used to treat pregnancy but after further research was proved very effective against skin cancers. Just because a drug fails to treat the intended disease doesn’t mean the drug doesn’t work.

But because of a lack of flexibility in the way clinical trials are designed means drugs that fail phase 2 studies are not able to be used to trial efficacy against other illnesses without restarting clinical trials from the beginning. But these drugs have already passed phase 1 studies. So a lot of probable cures are sitting in labs unused. A lot of babies are being thrown out with the bathwater.

These two changes alone would meaningfully impact the entire industry.


The reason for this regulation is ironically, to protect the consumer. If it weren’t so, drugs would experience a Moores Law-esque reduction down the cost curve. So if there was less regulation, drugs would be universally cheaper and there would be more and better drugs on the market since they would be cheaper to discover and make.

One theory says if you reduced the amount of compliance you’d have unstable compounds on the market. The other theory says you already have unstable compounds on the market so it isn’t working. There are even true horror stories of drugs approved that were catastrophically harmful.

It goes so far as to say the level of compliance hinders innovation because it prices them out. When the cost to commercialize a new drug is measured in billions, only already established and very large pharmaceutical companies like Lupin, Eli Lilly, Pfizer, Merck, Roche’s, Bristol-Myers Squibb and GlaxoSmithKline; can even participate in the drug creation process and only create drugs for widespread illnesses.

You’d have many more drugs with a higher efficacy since small pharmaceutical companies can then participate in R&D for niche diseases with smaller markets. Since they won’t need to rely on smash-hit drugs to succeed and disruption usually happens from the bottom anyway. A lot of cured diseases would then have medicines.

When you’re only trying to create medicines that address widespread diseases it means the amount of money that goes into research gets skewed. Why would you spend money researching an illness that effects 200,000 people when there are diseases that effect millions? And so a lot of low hanging fruit medical breakthroughs don’t happen as a result. Nobody sees the effect of this because how do you measure the impact of a disease that could have been cured but wasn’t?


But in an abstract way these are actually the good guys. For every big pharmaceutical that invests billions into R&D, there are hundreds of smaller predatorial pharmaceutical companies waiting for a patent to expire before flooding the market with their own branded medicines, who don’t participate in any R&D whatsoever.

This is what happens when a drug goes generic and why it becomes so much cheaper. These hundreds of smaller factories all start making versions of the same drug because the patent protection has lapsed and that competition brings down the price since they can charge less because they don’t have to recover any R&D expenditure.

These are usually the same factories who started as outsourcing before realising they could increase their profits by moving into manufacturing their own branded generics on expired patents.

It’s a bit like that scene in I Am Legend when Will Smith wakes up after the booby trap and the vicious dogs are waiting to attack. The only thing stopping the dogs from attacking is a beam of sunlight because they’ll burn if the sun touches them. As the sun is setting, the dogs become more and more anxious. When the sun finally goes down and the sunlight disappears the dogs attack tearing Will Smith to shreds. The sunlight represents a patent expiring. When the patent for Lipitor expired it wiped out 12 billion dollars from Pfizer’s bottom line because of generic substitutes.

What sometimes happens to stop this is when a patent is expiring and the company hasn’t found a dramatically new drug, they just make a small change to one of the drugs they already have and try to file a new patent. This is called “Evergreening” and it stops a drug from going generic but also keeps the price high which is why no one likes it when a company does it.

Could you imagine spending 20 years of your life building the company around a medicine and then it disappears and you have to start over? You’ve created culture and knowledge around this product watching it grow up and then you have to start from scratch. All your expertise is in one specific field and it’s very hard to suddenly transition into a new one.

It’s a strange industry in that way. Usually when a company creates a new product they’re allowed to earn money on it forever and improve on the product whenever they want. How long has Coca-Cola been around for? Or Mars Bars? But not in the drug world. A new product only has 20 years and if any improvements are made, it’s called a “me too” drug and slammed by critics.

Reinventing oneself is more difficult than it is made out to be. If you’re a company like Pfizer who have created Lipitor which in many ways is the perfect cholesterol reducing drug. Or GlaxoSmithKline who have created Ventolin which in many ways is the perfect asthma relief drug. The company becomes known for reducing cholesterol or relieving asthma. But it is difficult to medically improve a drug as effective as that.

It’s almost comical, the company which is known for reducing cholesterol or relieving asthma have lost the patent on the drug which does precisely that. It’s why often a company will follow up a drug with another one that does the same thing because the company is known for treating a particular illness.

When the patent expired for Lipitor or Ventolin which represents billions of dollars in revenue that would disappear. It’s reasonable to think a followup drug will earn a similar amount. Because the company knows what kind of marketing works for this kind of disease because they’ve been doing it for 20 years.

But this is actually a bad thing. Because it provides incentive to invest in only incremental improvements which means the opportunity cost is research into new areas, or potentially entirely new drugs treating entirely different diseases won’t happen. But this is an effect of having to hit performance or sales targets to have enough money to even be able to afford to commercialize new drugs in the first place. If the cost of commercialization went down, they wouldn’t need to do this.


Because it’s so expensive, something interesting you’ll notice is that funding sources are slim in the life sciences world. If you were a venture capitalist why would you put money into a company that will take 10 years to go to market whereas you could put money into the next cool tech web app and get a return much sooner. This is why public capital sources are so important to pharmaceuticals and why so many end up going public.

Whenever this much money is involved, the talent is also proportional. Talent in the pharmaceutical industry is off the charts. It takes hundreds or thousands of brilliant individuals to make a dent. Some of the smartest most brilliant people you will have never heard of. There’s nothing more noble than for someone with talent to apply it in a field that could positively impact everyone.

A scientist anywhere in the world who makes an important discovery is immediately going to be head hunted. One thing the industry severely lacks is good software developers working on meaningful problems, so a talented dev might want to consider Pfizer over Facebook.


What I think you’re going to see happen if the cost of creating new drugs doesn’t come down, you will have pharmaceuticals venture into orthogonally related businesses. For example you’ll start to see pharmaceuticals try and become medical practices or healthcare insurers or pharmacy chains to slowly transition out of the drug creation business. Everybody loses when that happens.

But here lies an interesting question, why haven’t pharmaceuticals done that from the beginning? The end result of a visit to the doctor is usually buying medicine. It makes sense for the people who create medicine to also sell it and prescribe it.

There’s always been this kind of standoff between drug creators and medicine practitioners. Pharmaceuticals versus doctors. But really they’re part of the same team. Both are necessary for a holistic approach to healthcare.

Each views the other as bad. Doctors see pharma as evil drug peddlers and pharma sees doctors as incompetently subjective. See 3 different doctors and you’ll get 3 different opinions. I know dozens of doctors who don’t even know how to check their email.

Something else people don’t realise is that behind every bad experience with a drug is actually a bad doctor prescribing it or a bad pharmacist not warning you about side effects. Most clinics and pharmacies actually still store all of their information in 1999 Microsoft Access databases. And others still use paper and filing cabinets.

It’s very surprising to people how much practicing medicine resembles just making a series of educated guesses. But guesses. A doctor doesn’t actually know what is wrong with a patient when they arrive. They in a sense deduce from a series of questions and differentials which follow Markov chains and mental pattern recognition.

And doctors are also human and very susceptible to peer pressure and medical fashion. It wasn’t until 1960 that germ theory had been developed and the medical community officially recognized that not washing their hands was actually killing people. Hundreds of thousands of people died because the medical community didn’t believe, despite huge amounts of evidence, that not washing their hands was a bad thing.

It comes from a fundamental difference in how each views patients. Doctors see patients as people but pharmaceuticals see patients as information since there is a layer of abstraction between them. The reason this goes so against the grain is because its dehumanising. It’s treating people like data points. People don’t like being treated like data.

But there is a line of thinking that says this might be the right way to do it. Because there are hard heuristics and tools. Software starts to emerge as a real healthcare tool. Which is what is happening now in its infancy.

The high cost also stops new companies from getting started. If you looked at the history of most large pharmaceuticals, you’ll notice they were all started a long time ago. Because of that you will start to see hybrids like mine emerge that start in software or pharmacy to generate revenues, venture into healthcare and only eventually into drug creation after you hit scale.


A lot of people say the world drug market is subsidized by the US industry which accounts for nearly half of the worlds pharmaceutical market. They’re actually right but for the wrong reasons that lie in history. A lot of welfare countries have artificially low prices for medicines. This is because they are heavily subsidised not because the drug company makes less on them.

The US for the last century was the global superpower with one of the highest quality of life and they invested heavily into preserving it. There is a reason the FDA is sanctioned by the US government. But over time and lax policies this investment accidentally gave a lot of firms monopolistic scale and is a function of nearly every new drug being in an environment where you can charge anything you want for it.

But it runs a bit deeper. In fact the world’s drug creation is subsidised by only a handful of firms. You can count them, there’s about 150. This is the really scary part. The one consistency is there is always going to be more people and people will always get sick. We’re living longer lives and are always going to need medicines.

So even though healthcare spending is high, as a function of how many people develop diseases, which is everyone, it’s next to nothing. And the burden of it is on the shoulders of only 150 companies, in the world. Talk about Atlas holding up the sky.