ASH abstracts are coming... beware of the headlines

It seems like we hear a lot about clinical trial results that are “significant.”  Yet, in many cases it feels like the outcomes of certain diseases really are not changing all that much.  ASH abstracts will be out in the next few weeks and it is always a time for news.  Unfortunately, much of the news is poorly reported because the language of science is not always the same as the language of the rest of the world.  Nowhere is that more important as the word "significant."  When the headlines scream "significant" it really helps to understand what is actually being communicated.

We often trumpet a study that has achieved a level of improvement to be considered “statistically significant” yet what that really means is that when two or more interventions were compared, the difference in outcome between the interventions is unlikely to have occurred by chance (or more accurately stated, likely to have occurred by random chance <5% of time should the study be repeated multiple times under similar circumstances).

The problem with this definition is that a small difference between interventions (say an improvement in response rate from 33% to 38% or improvement in survival from 11 months to 12 months) can be “statistically significant” if it is observed in a large enough population whereas most patients might say – “who cares if it is such a small difference.”   This is a key point, so I want to make sure it is clear.  If you see a 5% difference in a study population of 70 patients, you might agree that there is a good chance the difference is purely random.  On the other hand, if you see a 5% difference in a study population of 10,000 patients – chances are that is a real / reproducible difference.  In the latter case, we would call that “statistically significant” even if the patient says, “so what.”  Take a 50% difference in outcome however, and even if it is observed in a small population, it is a big enough difference to make you think it isn’t a random chance observation.

When we design studies we go through an exercise known as “powering the study” which enables us to project a difference between two interventions and then calculate how many patients we will need to study to enrolled to conclude that our difference is “statistically significant.”  If we project that a new treatment improves response rate from 20% to 80% that is a huge number and we need few patients to prove our point.  Similarly if we double the duration of response with a new treatment – that doesn’t take many patients either.

When the difference is small though, the studies have to get very large.  That is true when we already have very effective treatments (hodgkin’s disease) and you don’t have a ton of room for improvement (ie, can’t cure 130% of patients) or the incremental benefit is small (different hormone manipulations in breast cancer improving outcome by 1-2%).  One good clue to how meaningful a result is is simply to look at how many patients were enrolled.  If you have > 500 patients per arm, chances are the improvement is fairly modest.

Patients want “clinically meaningful” results such as “Dad survived 6 years instead of 6 months with his pancreatic cancer” or “everyone who takes the new drug feels better and responses are dramatically improved.”  Who could blame patients for wanting this.

Over the past 50 years most of our advances have fallen into the “incremental gain” category.  This is where we had huge studies to show that we could prolong pancreatic cancer survival by two weeks on average and this was trumpeted as “statistically significant” – yuck!  We’ve had a bunch of these recently in colon cancer.  Seethis link for a very good article about this.

Sadly, the route to approval of drugs requires “statistically significant” even if it is not “clinically significant.”  Of course, a new drug is going to be very expensive and if you have to take $90,000 of treatment to prolong life by several months, you might think twice if you were paying for it (provenge in prostate cancer).  The British have a system that measures “clinical significance” as part of their approval process.  I have to say that I can see some logic there – please look at this link for more. 

I am pleased that many of the experimental treatments in CLL fit the category of “clinically meaningful.”  It is important to note that randomized studies to measure the magnitude of difference have not been completed with ibrutinib, CAL-101/GS-1101, ABT-199, GA-101 and so forth – but they are underway.  Many thought leaders feel these agents will be both “clinically significant” and “statistically significant” to boot.  Hopefully we will gain broader access to these soon and patients will live longer, happier lives.

ASH abstracts are just around the corner.  You will probably hear a lot about “significant” results.  Pay close attention to the use of the terms “statistically significant” and “clinically significant” – they are different.  Look for how large the sample size is in the study.  Lymphoid studies tend to be smaller than breast / lung studies.  A big lymphoma study or CLL study might be >500 patients.  Keep in mind that you cannot define “statistically significant” unless you are comparing at least two groups – so they are either randomized studies or looking at subgroups within a larger study.

Hopefully we will have a lot of studies to discuss that really improve the quality of lives for patients with these disease.

Statistic vs real significance
drug cost vs efficacy

clonal evolution part 2

ASH plenary session link

I've written once previously about clonal evolution, but I think this is really an important topic and so I wanted to come back to it again.

There is a wonderful new technology called "next generation sequencing" that is turning cancer biology upside down.  The human genome project took 13 years, 6 billion dollars, and sequenced (ie measured every single piece of human DNA) the genome of four healthy individuals.  That is a lot of time and money.  You can now do the same (actually much more) amount of work in about 2-3 weeks at a cost getting closer and closer to $1000. 

When you can measure DNA at this level of depth at this cost you can start asking very important questions.  Take the following image:

This looks at an individual with CLL who had their cancer cells sequenced at 5 different timepoints in their disease.  I suspect this is true in lymphoma as well, but tissue is harder to get.  For now, assume this applies to both diseases.

At the first timepoint analyzed (a) there are already three "subclones" and a population of normal B cells.  The patient gets treated with chlorambucil and at timepoint (b) which is relapse following treatment, one clone has taken off as the major one (91%).  The patient then gets treated with FCR and overall the disease largely goes away.  Subclone 1 goes away forever.  Unfortunately subclone 4 which was only 1% of all the cells prior to FCR really takes off and becomes the clone that eventually causes the patient to get into trouble.

This highlights how the behavior of disease can change over time.  Different subclones may acquire different mutations (17p, 11q, BIRC3, SF3B1, NOTCH, etc.).  Though it may be lurking in background, it can become the predominant clone when exposed to therapies that eliminate the "easy disease."

I am not sure just yet if that makes an argument for how we treat patients, but I do think we aught to be looking to see how certain therapies affect patterns of resistance....

17p Deletion in CLL

17p is the genomic alteration in CLL that triggers the greatest concern in most patients.  It can have a tremendous impact on CLL prognosis and the FDA has recently extended approval to ibrutinib in this population (even without prior treatment) and the European equivalent of the FDA (the EMA) will do the same for idelalisib in combination with rituximab.  A lot of patients know that 17p deletions is one of the high risk markers in CLL  – but there are a lot of things to consider about CLL with 17p deletion before completely tearing your hair out.

When we say 17p deletion CLL, what we mean is that the short (petit) arm of chromosome 17 is missing.  You have 23 pairs of chromosomes (46 total) and as you get higher in the numbering, the chromosomes get smaller and smaller.  It is probably an excessive simplification to say that the biology of 17p is all about one particular protein called p53 – but for the time being that is most of the story.

P53 is affectionately called “the guardian of the genome.”  Every time I read about p53 I discover some new function of the protein that I didn’t know about before.  One of the most important though is that it will bind to DNA in a bunch of places and turn on / off the genes at those locations.  In this role it is known as a “transcription factor.”  Many of the proteins that are regulated by p53 have to do with cell survival or cell death.  When P53 decides it is time for a cell to die – very few things can stop that.  The most important signal that turns on p53 is DNA damage (hence – guardian of the genome). 

When DNA damage occurs the cells have a lot of repair mechanisms to try to fix the problem (including the ATM protein on chromosome 11q).   P53 will halt cell proliferation until that DNA damage is fixed.  Some DNA damage cannot be easily fixed and when that is the case, p53 triggers a cell death cascade called apoptosis (one of several ways that cells can die).

I mentioned above that you have two copies of every chromosome – so you ought to have two copies of P53.  We have been good at detecting absence of chromosome 17p for quite some time (via routine cytogenetics or FISH), but we have not always been very good at detecting p53 mutations which have been far more difficult to measure until recently.  With new sequencing technology, it is relatively easy to look for mutations and an increasing number of laboratories are offering that service. 

This is important because patients with 17P deletion are not the only individuals who have to be concerned about it.  About 30 percent of patients with abnormality in P53 have a mutation BUT NO DELETION.  Those have just as bad a prognosis but are not currently detected by FISH testing (nor SNP arrays which are one newer technology that is gaining popularity).   There is a strong association between loss of chromosome 17p on one chromosome and mutation of the other copy (about 85% of cases with 17P deletion will also have P53 mutation on the other chromosome).  

Another common misunderstanding has to do with “how many deleted cells does it take to call a patient 17p deleted?”  In other words, FISH will report the percentage of cells lacking one copy of 17p.  That can range from 1% to 100%.  In simple terms, the more abnormal cells, the worse.  For research purposes we say that 20% of cells lacking one copy of 17p calls that person “17p deleted.”  Some labs have lower thresholds (7%).  Occasionally I will hear from a patient that has 2% of cells with 17p deletion who is worried about their future.  By convention we would not group that patient into a 17p deletion category.

I think the 20% distinction is important – but gets more emphasis than it deserves.  We have prior posts talking about clonal evolution and this is a topic that is very important to understand (also covered in my "watch and wait" post.  If you have a small percentage of 17p deleted cells and you get chemotherapy that damages DNA – requiring p53 to transmit death signals – guess which cells are going to survive.  We know that one out of five patients will have a high risk molecular abnormality at relapse (11q/17p).  If we look hard enough we can see that it was often there to begin with – but below our typical levels of detection.  By giving therapy that removes the more sensitive cells, the resistant ones remain.  On the other hand, if you have a large number of 17p deleted cells, you are less likely to respond to chemotherapy in the first place.

The question becomes, what to do clinically when a patient has a 17p deletion.  There are not a lot of standard regimens that are particularly active when a patient has a high load of 17p deleted cells.  FCR and BR are not very effective.  Indeed, perhaps the most important clinical trial in this population right now is the frontline study of idelalisib in combination with rituximab.  It is available here and here (can be opened at any of these locations)

Campath (an antibody that does not damage DNA) can work well, but does not clear bulky lymph nodes which are common with 17p deletion.  High dose steroids can shift cells into the circulation where they can be removed by campath.  Rituxan also does not damage DNA both rituxan and campath combine well with high doses of steroids.

The new drugs CAL-101 (aka GS 1101), ibrutinib (aka PCI-32765), and ABT-199 (AKA GDC-0199) appear in preliminary reports to be quite active in 17p deleted CLL.  Multiple clinical trials are available for those drugs.  For untreated CLL with 17P, I think it is worth trying to get into this study

I have a particularly memorable patient who presented to my clinic with bad stage IV 17p deleted disease.  He had bulky nodes, WBC count of 200, platelets of 20k and hemoglobin of 8.  His FISH showed 100% 17p deleted.  Two cycles of FCR did nothing except get him transfused every few days.  I switched him to campath with rituximab and got his marrow into better shape but he still had bulky nodes.  He was young enough for transplant, but not eligible because he still had bulky nodes.  I sequentially gave him R-ESHAP, bendamustine rituxan, revlimid rituxan, ofatumumab all without much benefit.  I started him on CAL-101 and his disease melted away.  His disease control lasted nearly two years.

When a patient is young enough, they should definitely consider a stem cell transplant for 17p deleted disease.  The challenge though is that CLL more commonly affects patients too old for transplant.  The engineered T cells hold some promise for being active in this setting.

I also have a lower threshold for starting treatment in previously untreated CLL with 17p deletion (see "when to treat CLL").  Since those cells are likely to be resistant, I don’t see value in getting too far behind before getting started.   When I start, I might avoid FCR though some would argue it is still the right choice (NCCN lists this as first choice but I do not agree).  In Europe, you would typically get steroids with campath and I tend to think that is the right option.  Unfortunately, not enough sound data to tell us one regimen is better than another in this situation.  If a patient has access to ibrutinib in this setting that may be preferable.

Finally – one more biologic consideration.  Richter’s transformation is the name given to CLL that changes behavior and becomes a lot more aggressive – a different entity we call diffuse large B cell lymphoma.  It appears that p53 abnormalities are one of several key steps to getting to Richters (the other possibly being abnormalities in Myc or a protein that turns on Myc called NOTCH).  This is part of the reason Richter’s can be so difficult – it has intrinsic resistance to chemotherapy.

We are lucky to have a host of new drugs working through the system.  I will be very interested to see if drugs work out in this setting!

Thanks for reading - I also discuss this in a video done by Brian Koffman.  For anyone still interested, here is the link:   High risk CLL

CLL intro by the English

English CLL research has played a very important role in getting us to where we are today.  When they put out a 50 page document to familiarize patients with the disease - it is worth a look (as well as the funny way they spell Leukemia.

Here is a link to their PDF:
Patient Information: Chronic Lymphocytic Leukaemia CLL