Communists Fighting Cancer

There is a major cancer meeting at the end of May/early June in Chicago this year: the American Society for Clinical Oncology's annual meeting.  This is where scientists and clinicians get together to discuss all the leading directions in cancer research.  For those with a sufficiently technical understanding of the myriad details, it's kind of like a Roadmap for the Cure.  Some of the topics are about new research, others are about better treatment methods, others are simply about helping patients cope after they run out of options.

I don't expect my readership to have a strong technical understanding of cancer, so after the meeting I plan to write a post about what's new on the bleeding edge of cancer research - targeted at a a layperson's understanding.  Before I do I want to prepare you for what might come out of the meeting with a summary of where we are with cancer right now.

First, for the extremely uninitiated, I'd like to give a brief description of what cancer is, and what we think causes it.  You can go elsewhere to get the standard definition of cancer, though.  Instead, I want you to understand what cancer feels like from the inside.

Cancer: An Impressionist's Perspective

Large multi-cellular organisms are like communist dictatorships.  By which I mean the most extreme type of communism you can think of where not only is the dictator all-powerful, the citizens are extremely compliant.  If the Leader announces party purges, the random citizens who find themselves on the list of condemned will dutifully commit suicide.  Moreover, secret police constantly survey even the tiniest corner of the abandoned homes in small hamlets for intruders from outside the State.  And of course everyone is happy with this arrangement.  (All metaphors fall apart at some point.)

This kind of unity is important, because the State (body) is constantly under attack.  Outside, there are masses of insatiable creatures constantly trying to gain access to the State's rich supplies.  So long as everyone stays perfectly in line, there will be enough.  This is why I compare the body to a communist dictatorship.  Everyone works tirelessly to promote the community goal, not for their own benefit.  Anything that stop working explicitly for the community but tries to go it alone isn't just useless.  They're actively harmful and must be destroyed.

You can think of cancer as like a resistance movement.  A group of cells decide to stop acting in concert with the communist dictatorship and do their own thing.  In this way, they're similar to the invaders outside, but it's difficult for the body to distinguish between obedient cells and rebel scum.  It's not like they're going about wearing badges.  "I'm an insurgent, kill me now!"  Also, they don't have to get past the border wall.  They're already inside and able to take advantage of the same rich resource supply the state relies on.  The cancer insurgency, therefore, has some natural advantages over invading forces from the outside.  It also has some disadvantages, which we'll talk about below.

Like any good resistance movement you need multiple levels of organization to succeed even a little bit.  Mammals have been fighting cancer-style resistance movements for millions of years, so they're really experienced at preventing and fighting against cancer.  They know all cancer's tricks and have strategies for defeating each one.  To survive, cancer insurgencies have to get creative.

Step 1: Don't Die

Like any totalitarian state, the body can undertake grand building projects at low cost (only the cost of life, as it were, and life is cheap if you're a body or a ruthless dictator).

A great example is finger development. While a fetus is developing, instead of building each finger out from the wrist, the body grows a whole hand paddle, then carves five fingers out of the solid pad of skin. It does this by telling the cells between fingers to kill themselves, which they dutifully do. You can see the remnants of this process if you stretch your thumb and forefinger into an 'L' shape. There's a little wing of skin still there. Some people lack the signals the body uses to sculpt the fingers and they end up with webbed fingers (if you're faint of heart, you may want to skip this link).

Remember, everything that happens inside your body has to be pre-programmed.  So all these cells that kill themselves off so you can have separate fingers have the suicide instructions built into their DNA.  They're like tiny computers and when the signal to self-destruct comes along they obey every time.  Every cell has this self-destruct mechanism built in, and there are multiple redundant systems so if one fails the others will take over.

In cancer, the cell learns to overcome this programmed cell death.  Sometimes it can get by only disabling a few cell death signals, not all at once.  Once it makes this change it never has to die.  But it's still only one cell...

Step 2: Recruitment

Any good rebellion is going to need rebel fighters.  Fortunately, if you're a cell you don't have to convince your neighbors to join you.  You can just make copies of yourself.  Have enough children, and your descendants will fight for you.

But the authorities are wise to this approach.  There are strict instructions in every cell type, and most cells aren't authorized to reproduce.  This might seem counter intuitive.  Don't you shed billions of skin cells every day?  They're a major component of house dust, with old dead skin cells shedding all the time.  How can most cells not reproduce, but you're constantly making new cells?

There are two different types of 'authorization to reproduce'.  Down at the base of your skin there are cells we call adult stem cells.  They're not like the famous stem cells that can become anything.  They're just a special kind of skin cell that is allowed to reproduce - very slowly.  They can divide in one of two ways.  Say you get a paper cut and the skin cells need to grow back together as part of the healing process.  The adult stem cells divide, producing exact copies of themselves, and fill in the cut.

Normal skin growth is a little different. This time, the adult stem cells divide horizontally, not vertically, producing one cell that stays at the base of the skin and another that starts its journey toward the surface.  This happens maybe once a week for each of these cells.  The cell that leaves the base for the surface is different from the adult stem cell.  It's now a progenitor cell, which can no longer divide forever; it has a fixed number of times it can divide.

But it traded the ability to divide forever for the ability to divide very fast.  It makes lots of copies of itself - about once every eight hours or so - and creates a nice thick layer of skin.  Once it reaches its maximum number of divisions, it stops dividing and dies.  This layer of dead cells is what your skin, hair, and fingernails are made of.  As cells on the outside are sloughed off, they're replaced by the next-oldest layer, on a continuous conveyor belt from the base of the skin up to the surface.
If you're trying to start a cancer revolution, this mechanism (presented here for the skin, but it's present in more complicated forms in other tissues as well) is where you'll want to target your recruitment efforts.  Each type of division presents different strategic opportunities.  For example, you could try subverting the progenitor cell's restriction on number of divisions.  Then you could continue growing forever.

This type of mutation is primarily what we're targeting when we use chemotherapy or radiation.  One type of programmed cell death looks at how badly the DNA is damaged and causes the cell to self-destruct if there's too much damage.  Chemotherapy causes massive amounts of DNA damage to any cell that's in the process of dividing when you take it.  That includes progenitor cells everywhere, not just the cancerous ones.  So your hair falls out as the skin progenitors die off, and you get violently ill because the cells lining your gut die off.  You also have to be careful not to get sick, because most of your immune system cells get wiped out, so you can't fight off even small infections.  It's a difficult but effective treatment.

The problem is you have to get every last cancer cell to succeed.  If even one rebel cell survives it can grow out and become a problem again.  If you got chemo, but it wasn't strong enough - or didn't last long enough because a few of the cells were in between divisions when it hit - it might take a month couple months to go from the handful of cells you missed to a couple billion again.  Maybe longer, depending on how fast the cells grow.  There's no way to see fifty scattered cells on an MRI right after chemo, so your doctor will say they 'think' they got it all because the tumor disappeared.  But some cells might be lurking in the background, waiting to grow out into a new tumor.  You'll come back in a couple months for a follow-up scan to see if that's the case.  If the last round of chemo wasn't enough, we might need to hit you with a higher/harder dose to get it all next time.  This is part of where you get this imagery of a 'fight' against cancer.  It's a war and you have to root out every last insurgent or they'll just keep coming back.  The patient has to fight, enduring multiple rounds of chemo until every last rebel cell is eliminated.

A second recruitment method a cancer insurgency might use would be to target the adult stem cells.  They only divide once a week, so they could be changed to divide all the time.  This is difficult, because one of the safeguards your system has against rapid cell division is any cell that divides too quickly is pre-programmed to kill itself.  You could subvert this system and end up with a similar result as when you converted a progenitor cell.  This would also be susceptible to chemotherapy.

A better strategy would be to combine the progenitor system with the stem cell system.  Instead of making stem cells that rapidly divide directly, the stem cells could produce progenitor cells capable of living forever without a restricted number of divisions.  The stem cells still only divide once a week, but each time they divide they produce a new rapidly-dividing cancer cell.  In this case, if you hit the cancer with chemotherapy you'll wipe out all the cancer progenitor cells, but miss most of the stem cells.  Sure, they divide once a week or so, but they don't divide often enough to catch all of them with reasonable doses of chemo.  To get rid of this kind of cancer you'd have to hit someone with so much chemotherapy it would also eliminate all the adult stem cells throughout their body.  That means skin, immune, gut, liver, kidney, and every other cell type would never be able to grow back.

You'd be dead.

The problem your doctor has is they can't tell the difference between the kind of cancer that uses this stem cell-progenitor cell system and one that's simply rapidly dividing.  Many types of cancers can be treated with chemo and never come back.  It sucks to go through, but you survive and continue on with your life cancer-free for years.

Or it comes back after chemo.  When you get the second round of chemo, you don't know whether you're fighting the last few cells that you accidentally missed last time, or whether you're fighting cells that no amount of chemo will kill.

Step 3: Infrastructure Development

After it figures out how to stay alive and grow an army cancer is faced with a new problem.  The totalitarian state won't give it more resources than what's absolutely necessary for survival.  In other words: it needs blood vessels.  Without new blood vessel growth, the fledgling cancer army can't get much larger than the head of a pin.  To get nutrients, it needs to learn how to send signals out to grow new blood vessels.  Until they learn to do this, cancer cells can't form tumors.

Often cancers will send these signals out so much that the blood vessel cells are also rapidly dividing.  They're still normal cells, they're just responding to abnormal signals.  This presents an interesting opportunity for cancer researchers.  If you can target the blood vessel cells, not just the cancer cells, you could control the cancer growth.

Some cancer researchers were looking at cancer cells that get around programmed cell death - the kind that is activated when chemotherapy tells a rapidly-dividing cell to kill itself.  These cells shouldn't respond to chemotherapy, because they developed mutations that undermine chemo.  But noticed sometimes these cancers were still affected by chemo.  This didn't make sense.  Why did these cells respond to a treatment they'd mutated to avoid?  The answer was in those rapidly-dividing blood vessel cells.  Since the blood vessel cells are normal - not mutated - and rapidly-dividing the chemo told them to kill themselves.  This cut off the supply chains and caused the tumor to starve and collapse.

Step 4: Evade the secret police

When planning your insurgency, everything you do has to be pre-programmed into the cell.  This is a major restriction.  You can't just write a new program wholesale, like "grow out of control".  You have to take existing programs for controlled growth and modify them to do what you want.  The way cancer changes the programming is through mutations in DNA.  But there's no program to 'make cancerous mutations', so it has to rely on an existing program: random mutations.  Mutations happen randomly throughout life.  They happen in the background, not just because you're exposed to some chemical that's known in the state of California to cause cancer.  The sun, background radiation, or just crazy random happenstance all cause mutation.  That's right, your DNA sometimes mutates for no reason at all.  You can't stop it from happening unless you're frozen down to absolute zero.  (At which point cancer isn't your biggest problem.)

There are some things you can do as an individual to increase the rate of mutations.  For example, you could take up smoking and expose your lungs to a constant barrage of mutation-causing chemicals.  There are also things your cancer can do to increase the mutation rate.  I was about to go through some, but that's beyond the scope of this post.

Whether the mutations come from the outside or are driven by the cancer itself, they can help the cancer when they make changes to normal cell programming.  Sometimes that means modifying a protein (think of proteins like interacting nano-machines that work together in the cell factory to make life happen).  The modification might remove the 'off switch', or make it not turn on, or make it go somewhere it isn't supposed to.  Each change makes the protein look very slightly different from normal.

I won't go into how it all happens here, but one of the body's best defense mechanisms against cancer is the immune system's ability to recognize tiny changes.  This ability is why people who get organ transplants have to take immune suppressors for the rest of their lives.  No matter how close a match it is, your immune system can tell the difference between the kidney of your closest sibling and your own kidney.  This applies to any changes cancer cells make to proteins in order to modify their programming.  Immune cells that see these altered slightly proteins will treat the cell that harbors them with extreme hostility.  They circulate around in the blood constantly looking for offenders.  Nobody escapes their notice.

You might think a cancer insurgency would be safe to just make mutations to stuff inside the cell.  If it doesn't make any changes on the surface these cells that circulate through the blood won't be able to get at it, right?  Wrong.  There's a mechanism in every cell that randomly samples everything made inside the cell.  It chops up pieces of every type of protein and displays those pieces on the surface of the cell.  It's the ultimate snitch.

What do you do with snitches?  You get rid of them.  The cancer insurgency decides to stop the snitch protein from going the the surface and displaying all the changes inside the cell.  Except - and here's where you just have to admire the genius of this kind of system - there's another kind of immune system cell that goes around looking for the snitch on the surface of every cell.  If a cell doesn't have snitches - or if the snitches don't look familiar - it gets slaughtered.

So how does a cancer insurgency evade the secret police?  One simple method is to put up a 'go away' sign.  It can tell all those immune system cells to go look elsewhere.  Another strategy is to co-opt the immune system.  It convinces the police there's nothing wrong with what they're seeing.  It can even get the police to lobby for additional resources - like new blood vessels.

Quelling the Insurgency

Targeting this last approach is behind some of the newer immune-system related treatments, called 'checkpoint inhibitors'.  Basically, it targets the reprogramming step cancers use to co-opt the immune system 'police' cells, preventing them from being turned.  Suddenly the vast tumor conspiracy becomes visible to the immune system, which begins attacking and destroying it.

Another approach along this line has been confusingly called 'cancer vaccines' in the past, but is now commonly referred to as CAR-T (although that oversimplifies everything, and they're using more than just T-cells these days).  The idea is to take some of the police cells out of the body, send them to a lab, and train them on dead cancer cells from the patient.  This is like giving a piece of clothing to a blood hound, because the victim is nowhere in sight to try and reason with the hound while it's picking up the scent.  The tumor can't reprogram the police if they're at a special training session far away.  After the cells have been trained, they're put back into the patient where they attack any cancer cells that get out of line.  Since the cells are long-lived cells, they'll still be around years later - hopefully providing a lasting 'cure'.

These two approaches are all the rage right now, and have been providing some promising clinical benefits.  CAR-T is probably the most expensive and technically challenging method to date.

There's another approach you could take that's more specific than any other so far.  We call it 'precision medicine'.  It takes advantage of all the work we've been doing over nearly four decades of cancer research, in which we've been identifying each of the different genes cancer cells can reprogram to aid their insurgency efforts.  This includes dozens, if not hundreds, of genes and multiple different approaches for each gene.  Some of these genes you've heard of, like BRCA1 and BRCA2.  Others you've never heard of, like TRK3.  The name isn't important.  The important point is that after decades of research we have a rich understanding of the specific ways cancer can co-opt the cellular machinery.

The problem is that there are so many different ways to choose from.  And since mutations are random, you can't just target all cancers with a treatment against BRCA1 mutations.  Most patients wouldn't be affected by a treatment against BRCA1.  Looking at only one mutation necessarily only targets a small percent of cancers.  The obvious solution to this problem is to create treatments against all the different genetic mutations.  Then you could just test each patient to figure out which genes their cancer is taking advantage of and treat them with drugs that target those genes.

That's a great idea, but it requires that you test for hundreds of genes.  In the past, that would require more than a whole tumor of tissue to get through a small fraction of the genes, testing them one-by-one.  However, a number of new technologies, most notably Next-Gen Sequencing (or NGS), allow us to test hundreds or thousands of genes at once.  As a result, some new treatments are coming out that work on exactly this principle: test for genetic abnormalities and treat patients based on the genes specific to their cancer.

One vision for the future of cancer treatment is that after your initial diagnosis your doctor sends some tissue off to get sequenced.  When it comes back, instead of trying chemo or some other caustic therapy, he gives you two or three pills that target the exact mechanisms your tumor needs to grow and survive.  If we could hit more than one mechanism at a time, it would be similar to how farmers use multiple different herbicides on their fields simultaneously.  Even if the cancer were able to mutate around one attack it can't handle three or more at the same time.  This is putting natural selection to work for us.

How far away is this therapy?  Well, now that we can detect large panels of mutations at the same time we need to develop treatments against each target gene.  That step has already begun.  I've worked on some pretty spectacular trials where near-terminal patients were treated and their cancers shrank down to nothing with almost no side effects.  The problem is that each drug is only going to cover a small percentage of patients.  Some less than 1%, some more.  It's unlikely we'll see any that are over 50%, or even much more than 20-30% (if we're lucky).  Many will overlap, so we need to get more than just a cumulative '100%' coverage.  That means we'll need dozens, maybe even hundreds of new cancer drugs, each targeting a small number of cancers.

We have plenty of candidate targets, and there are already dozens of clinical trials attacking many of these genes.  Still, the work will take years before we have multiple genetically-based treatments for every cancer.  Along the way, we'll slowly go from a few lucky individuals whose cancer can be treated this way, to a small but significant minority of patients, on up until most patients can be treated this way.  In twenty years or so the next generation's cancer treatment will look completely different from how it looks today.

Their doctor will collect a biopsy and send it off for sequencing.  Once it comes back, they'll be prescribed a specifically tailored cocktail of medicines, which will attack the cancer from different angles all at once.  The tumor will disappear, never to return, all with minimal side-effects.

At least, that's the dream.  Along the way, we'll likely see advances with checkpoint inhibitors, CAR-T cells, and a host of other approaches.  Each advance will make treatment with cancer a little more tolerable and give a little better life expectancy.  The next generation may never understand why cancer was so terrifying to us.  The future creeps in so slowly.  We may not be able to celebrate the specific day when we have the cure in hand, but it will come.  People say, "Each step brings us closer to the cure."  I would say each step cures someone new.  And the plurality of cures will spell the ultimate demise of the cancer insurgency.

If you have additional questions about cancer, ask in the comments.  Cancer is complex, but I'll try to make it accessible.

UPDATE: From ASCO, I wrote about new combination therapies, how biomarkers are making current therapies work better, and a summary of how NGS will change everything.


Popular posts from this blog

Reverse Engineering Life

Cancer update: precision oncology comes into its own