Tuesday, November 5, 2013

11q deletion CLL

Most patients who make the effort to learn about FISH in CLL know that 17P is bad and 13Q is generally favorable.  Trisomy 12 segregates based upon Notch mutations (if you can find a way to test for it) .  Somewhere between the badness of 17P and the middle of the rode trisomy 12 lies 11Q.

CLL with 11Q deletion has a unique personality.  It is notorious for having disproportionately bulky lymph nodes compared to the elevation of white blood cells (though WBC elevations are common).  Men with 11Q are considerably more common than women, and patients are often younger than the typical 72 year old average.  Furthermore the substantial majority also have unmutated B-cell receptors (unfavorable).  Patients often have need for initial therapy more quickly than others following diagnosis and patients with 11Q have shorter survival than many others.  This is in part because 11Q is often thought of as a “chemotherapy resistance” marker and I wanted to go into a deeper biology of this marker as I have for 13Q, trisomy 12, and 17P.

Once again, recent biologic insight has dramatically reshaped our thinking about 11Q.  We have previously addressed “clonal evolution” where CLL is a mixture of different “subclones” that jockey for dominance.  In the past it was observed that patients with greater than 25% of cells on FISH had worse prognosis than patients with less.  In light of clonal evolution, I tend to just think it is a continuous spectrum and if you have an 11Q clone it is likely to become dominant over time in most cases.  At any given time point, the less you have the better.  Like all other high risk genomic alterations, 11q deletion increases over time (10-15% at initial diagnosis but as high as 30% in relapsed / refractory).

Classically, 11Q deletion was thought to be all about a single protein called ATM (for ataxia telangiectasia mutated – the gene that causes a rare genetic syndrome Ataxia Telangectasia).  The gene for ATM is truly massive.  The DNA that encodes the gene (like most all genes) is broken up into a bunch of smaller pieces (exons) that all get glued together into one sequence (mRNA) before the cell can use the template to synthesize the protein.  The protein is very large and has quite a few different functions.  Consequently older techniques of measuring mutations in DNA (sequencing) couldn’t really address whether an individual patient had a mutation in the gene or not.  Furthermore, we didn’t really know the functional significance of many of the mutations because they were scattered throughout the entire gene.

Newer sequencing technologies allow us to measure with considerable precision BOTH the mutations AND the fraction of cells that have them.  It should therefore not come as a surprise that we have found an important number of cases where ATM can be mutated whether there is an 11Q deletion or not.  It is very similar to the P53 story on 17P.  You can therefore have deletion with or without mutation and vice versa.  It appears that having BOTH deletion of ATM on one copy of the chromosome and mutation on the other copy of the chromosome is worse than only having the deletion.

Whether it is deleted or mutated or both (you do have two copies of the gene), ATM plays an important role in detecting DNA damage, passing that information along to P53, and triggering an arrest in cell division and cell death.  Without ATM in place, you don't have the same response to DNA damage.  Drugs like fludarabine, cyclophosphamide, bendamustine, chlorambucil all require an intact DNA damage response for their efficacy. 

While that may seem complicated – it gets a lot worse.  I wrote quite a bit about the deletion size in cases of 13Q deletion where there are type I, type IIA, and type IIB chromosome changes that may have different prognostic significance.  Not all losses of DNA are equal to one another.  Bigger deletion sizes may include other important genes.  Indeed, the far end of the 11Q deletion variably affects a protein I’ve written about previously that is extremely important – BIRC3.  It has been shown recently that some 11Q’s have loss of BIRC3 and some don’t.  While we don’t directly know how important that is just yet, BIRC3 is a big deal as having a mutation in that protein is about as bad as 17P.  I think we will be hearing more and more about BIRC3 in the coming years as we look at it more frequently.  Finally, there seems to be an association between mutations in the protein called SF3B1 and cases with 11Q.  SF3B1 is not in the region of the deletion - BUT - there seems to be a relationship between this mutation and 11Q (sort of like NOTCH and trisomy 12).  SF3B1 is a mutation that confers resistance to fludarabine and possibly other chemotherapies.  Why this associates with 11Q really isn't known just yet but it is an important observation.  Patients with SF3B1 mutations often need therapy earlier on and don't do as well in the long term.

Wikipedia has a short description of BIRC3 here.  This pathway is also quite important as it governs the "execution pathway."  Once a cell has made the decision to die, it uses a series of enzymes called caspases.  BIRC3 is involved in governing that pathway.  BIRC3 abnormalities appear to be particularly bad thing to get.  Some data sets make it look just about as bad as having 17P deletion.  I think we are only at the start of understanding the interaction between ATM and BIRC3 in terms of deletion size, inclusiveness, exclusiveness etc.

Terry Hamblin did a very good job explaining the role of alkylating agents in CLL with del 11q, you should definitely see his discussion of the topic - link here.  In short, if you are going to give fludarabine based regimen to a CLL patient with 11q minus, many experts think you should include cyclophosphamide (ie. FCR more than FR) if the patient can tolerate it.  Some have even gone farther and asked if you need the "C" in FCR if you don't have an 11q deletion.  Most of this was based upon data generated when fludarabine was compared to fludarabine / cyclophosphamide (and rituximab wasn't in the mix).  Newer publications have suggested that the combination of FCR overcomes the negative prognostic implications of del(11q).  They definitely respond well to the regimen, but earlier relapse remains an issue.

Some are anticipating the results of the big German study comparing FCR to BR in the frontline setting.  These results may come as early as this December at ASH.  I am most interested in the subgroup of patients with 11q deletion in this study because of the question marks about alkylating drugs (chlorambucil, cyclophosphamide, bendamustine) in this patient group.

New drugs like ibrutinib and idelalisib do not have a long track record in subgroups like 11Q.  In our New England Journal of Medicine article, we saw that patients with 11Q OR 17P didn't do as well as patients without either abnormality.  The still did very well on this drug - but I expect that 11Q will remain problematic under most circumstances since it is a marker of genomic instability which is NEVER a good thing in cancer.  My guess is that earlier utilization of these targeted agents in smart combinations with other drugs may be best strategy (I've written about that here).

Anyhow, that is probably enough for now.
Thanks for reading.