What is Tumor Lysis Syndrome (TLS)?

Tumor Lysis Syndrome (TLS) is the term applied to the death of cancer cells at a rate that exceeds the body's abilities to deal with the consequences.  While we normally think of a "tumor" as a solid mass, this syndrome is probably more common in the lymphomas and leukemias then it is in cancers of the lung, breast, colon, and prostate.

While we might think that killing cancer cells is a good thing - killing too many at once can be a disaster.  In general this is a sufficiently uncommon problem that most patients do not need to worry about it - but for patients receiving treatment for any lymphoma or leukemia it is reasonable to ask your doctor if it is much of a concern and what to watch for.  I find that most patients are pretty unfamiliar with TLS so I though a brief description might be useful for patients who want to understand more.

In human physiology, cells have an extremely important boundary called the "plasma membrane."  It essentially is the zip-lock bag that keeps the insides in and the outsides out.  It is pretty remarkable because the insides of cells are a lot different in composition than the environment outside of the cells.  Cells spend a lot of energy keeping things that way.

If you consider something simple like potassium - there is a lot of it inside of cells and very little outside of cells.  Sodium is the exact opposite - lots on the outside and not much inside.  This separation of ions is important because it serves as the basis of a lot of electrical impulses in a wide variety of cells.

Under appropriate stimuli, you get a sudden reversal of these ion distributions.  Sodium flows in through specific protein channels and potassium can flow out just as quickly (so that we don't spark).  This creates an electrical current that can serve a ton of important physiologic needs.  In nerve tissues, it is what helps send nerve impulses all the way from your toe to your brain.  In cardiac muscle, it is what helps coordinate your heart beat.  In B-cells it helps the b-cell identify that the b cell receptor just found the virus it was supposed to fight off.

Just like all aspects of human physiology, things can go wrong.  Push yourself too hard in a sporting event and you overwhelm your skeletal muscles ability to keep things in balance - you get a cramp.  In cancer medicine, we occasionally encounter "tumor lysis syndrome" where we kill off a bunch of cancer cells all at once - and the insides of the cells ends up outside the cells causing a variety of problems.  Do this slowly enough and the kidneys and the liver clean up the mess.  Do it too quickly and those organs can't keep up.

Since we already talked about sodium and potassium we might as well start there.  When too many cancer cells die at once (and the inside / outside barrier breaks down), you can sometimes see sudden dangerous elevations of potassium.  While potassium is normally a good thing too much potassium is very dangerous.  As I mentioned above, potassium balance helps your heart beat right.  While the body is pretty good at adjusting to slow changes, if you suddenly raise your potassium it causes the electrical impulses of the heart to become ineffective and in extreme cases can cause it to stop.  Fortunately we can often bring it back down by giving lots of IV fluids, diuretics, insulin, sugar, and even something that helps you poop it out.  Administering calcium can help reset the balance as well.  Sometimes people even need temporary dialysis to get rid of the extra potassium.

Potassium isn't the only problem though.  All that cellular DNA has to go somewhere.  It gets broken down into a compound known as uric acid.  Folks with gout probably know a little about uric acid because it can sometimes form crystals in joints like the big toe or the knee which can be extremely painful.  In tumor lysis syndrome, sometimes those crystals form in the kidney and can cause significant organ dysfunction.

Once again, there are lots of things we can do to prevent this like IV fluids (to flush the through), change the pH of the urine so that it is harder for crystals to form, administer a medication (allopurinol) that prevents the formation of uric acid, or even a super expensive medication known as rasburicase which can break it down into smaller molecules super fast.

Phosphorus can also cause problems as it oozes out of dying cells.  Phosphorus likes to bind onto calcium and make calcium-phosphate.  In bad tumor lysis syndrome that can cause your calcium to start to fall (another problem with muscle tissue), or it can even form crystals in blood vessels and block blood flow to the skin.  That would typically be in a pretty extreme case of tumor lysis.

So while killing cancer cells is a good thing - you don't want to kill them all too fast.

Tumor lysis syndrome is most common in the really fast growing blood cancers like acute leukemia (not chronic) or the intermediate / high grade lymphomas like DLBCL or Burkitts.  See my post on the difference between these.  Sometimes these cancers grow so fast that you see an "intrinsic" tumor lysis syndrome even without any sort of treatment.  The cells are proliferating so quickly that the ones dying around the edges can cause all of the problems above.

It is pretty uncommon in the slow growing lymphomas and infrequent in CLL though more of a problem with some of the very effective treatments we have now (including FCR and some research drugs).

As in most things medical we need to be able to quantify TLS.  Therefore there are two categories of TLS - laboratory TLS and clinical TLS.  In the former, there may just be some mild asymptomatic laboratory changes.  In clinical TLS you have laboratory TLS and either kidney dysfunction, heart arrhythmias, or seizures.  Clinical TLS is graded on a scale from 0-5 with 0 being mild kidney abnormality and 5 being fatal.

In 8 years of practicing oncology I cannot recall a patient who either had a seizure or died from TLS (read: uncommon), though I have certainly seen bad electrolyte changes, renal function shutting down (though it typically comes back), and cardiac rhythm abnormalities.  That said, I am aware of a number of my colleagues who have taken care of a case that got that bad.

Anyhow - I am struggling with how to end this post.  I do not want to alarm patients unnecessarily, but I was thinking about tumor lysis syndrome and figured a good number of patients may not be aware of what it is.  This is meant more for education than any specific warning.  Hopefully I have not added to the anxieties of having one of these diseases.

Once again - thanks for reading!

p53 matters in DLBCL too.

I've written about 17p deletion in CLL previously.  You may recall that this is the chromosomal home of TP53 - the gene which encodes the p53 protein- one of the most important determinants of chemotherapy success.  P53 is mutated with lower frequency in many of the lymphoid cancers than it is in solid tumors (such as pancreatic, esophageal, etc) which may partially explain the success we enjoy in treating diseases like DLBCL.

A recent paper has evaluated the impact of p53 mutations on survival of patients with DLBCL treated with R-CHOP chemotherapy.  They found a mutation in about 1/5 patients.  It was important!  Those without a mutation survived about twice as long as those in whom it was not mutated.

I've attached a link to an editorial / summary I wrote on behalf of Clinical Oncology News for the interested reader:  p53 Is a biomarker in patients with DLBCL on R-CHOP.  Below is a copy of the text.  It is written for an audience of oncologists so the writing might be a little more technical than what I usually put here in the blog.  Hopefully it is still worth the read.


Personalized medicine is the goal of detecting unique characteristics of an individual’s cancer and appropriately modifying therapeutic interventions to maximize efficacy and minimize side effects. A growing number of molecular diagnostics offer multiplex analysis of many oncologic targets bundled within one commercial assay. With the explosive progress in genome sequencing technology, a thorough understanding of molecular biomarkers is imperative if the goal of personalized medicine is to be reached.

p53 is one of the most important molecular markers in cancer. Known as the “guardian of the genome,” p53 determines cell fate in response to a variety of cellular stresses by regulating transcription of important proteins resulting in cell cycle arrest or activation of pro-apoptotic machinery. Loss of p53 activity is therefore a common mechanism by which cancer cells avoid cell death in response to chemotherapy.

The article by Xu-Monette et al highlights the importance and complexity of p53 analysis in patients with DLBCL. Despite histopathologic similarities, patients with DLBCL can be clustered into distinct subgroups based on gene expression profiling. Beyond RNA expression differences, DNA mutational analysis adds further insight into the prognosis of these patients.

This study evaluates a large multi-institutional cohort of patients with de novo DLBCL (excluding patients with transformed disease), treated in uniform manner according to p53 status, using a variety of molecular techniques including sequencing, expression profiling, fluorescence in situ hybridization and immunohistochemistry.

p53 mutation is shown to be an adverse molecular finding in the 21.9% of patients with abnormalities. Overall survival in patients with wild-type p53 was nearly twice as great as it was for those patients with mutated p53, with similar effect on PFS. This effect was independent of germinal center (GC) or activated B-cell subtype, which differs from the prior analysis of DLBCL patients treated with CHOP alone (pre-rituximab) where the effect was limited to patients with GC subtype DLBCL. This highlights the importance of re-evaluation of prognostic markers as standards of care change.

One concern is the incredible complexity of p53 alterations. This study highlights the multitude of ways p53 can be altered in gene sequencing. Although there are hotspots for recurring abnormalities, not all mutations are created equally. Some mutations may not affect amino acid sequence, whereas others cause an amino acid substitution or premature termination of the coding sequence.

It is tempting to consider p53 changes in a binary “yes/no” manner but that probably oversimplifies the biology. As personalized diagnostics emerge, the interpretation of such abnormalities may be as important as the detection of the change in the first place.

Identification of a high-risk cohort may enable therapeutic intensification, but such trials are difficult to design and execute. As the molecular taxonomy of cancer advances more quickly than our ability to know what to do with the data, clinical research remains a vital link in advancing the care of these patients.

Richter's Syndrome / Histologic Transformation

Patients with CLL or indolent NHL occasionally experience a significant clinical change in their disease where it becomes a lot more aggressive.  When this happens, the formerly "slow growing" cancer becomes a lot more nasty and in many cases the prognosis gets a lot worse.  In a number of publications, NHL, CLL) the average survival when this happens is about a year.  A number of those are older articles (retrospective / in pre-rituximab era) which may have been confounded by patient selection bias.  My own impression is that many patients do quite a bit better but that is at least what the literature reports.

In CLL/SLL this is called "Richter's Transformation (RT)" while in the indolent NHL's this is called, "Histologic Transformation (HT)."  Sometimes docs jumble these terms and call it "Richter's Syndrome" or "transformation" regardless of which disease it started out as.  There is a different discussion about what we call Grade 3 follicular lymphoma.  Sometimes these can be confused by the patient.  I will save the discussion of Grade 3 until an upcoming post.  In follicular lymphoma HT occurs at a rate of about 3%/year.  While that is a pretty small number, it is cumulative so by 10 years it may be as high as 30%. In CLL the rate appears to be a fair bit lower so that the cumulative risk is only about 10-15%.

The best clue that a patient has undergone RT/HT is when the disease acquires a bad attitude.  Instead of just involving blood and lymph nodes, you see it in new places like liver, lung, intestine, bone nodules, sometimes even brain.  Patients might experience increasing fevers, night sweats, weight loss.  Laboratory changes are notable for a significant rise in a blood marker known as LDH (we are not talking about subtle changes, but 2-4x higher).  If you get a PET scan (which measures metabolic activity of tissues), you might get one spot which is disproportionately "hot."

Traditional risk factors for developing this in CLL include an increasing number of prior therapies, CLL diagnosis at a younger age (longer exposure to risk), and more advanced disease.  A number of newer studies show that pre-existing NOTCH mutations, "stereotyped B-cell receptors (a topic for a future post)," 17p deletions etc. also increase the risk.  In indolent NHL, risk factors include the diagnosis of grade III follicular NHL, advanced disease, high flipi scores, and several lab variables (LDH, B2 microglobulin).

Under the microscope, the new disease most commonly resembles the "intermediate grade" Diffuse Large B Cell Lymphoma.  Less commonly it can look like Hodgkin's Disease, and extremely rarely it may look like Burkitt's or Lymphoblastic Lymphoma.  In any case, it goes from "indolent" to aggressive, or even the highly aggressive.

Because it is so easy to get samples of cancer cells from patients with CLL (blood draw), we know a lot more about transformation in CLL than we do in low grade lymphoma.  It is probably worth while therefore writing about what occurs in CLL and then highlighting the differences that we know about in NHL.

In CLL there are two main and one uncommon way of experiencing RT.  The most common way (80%)  is for the dominant CLL clone to acquire more and more genomic mutations over time.  Typically these involve several important genes including p53, Myc, and NOTCH.  The second most common way (20%) is for a patient with CLL to "spontaneously" develop an entirely new diffuse large B cell lymphoma that is clonally unrelated to the original CLL. You might think no patient should ever have such bad luck, but in a prior post about 13q, I detailed how some genomic deletions can predisopose to lymphoid malignancies.  Some patients who develop CLL may in fact be predisoposed to the spontaneous development of DLBCL.

The difference is significant.  In the first case, you have a highly resistant clone - often with a p53 mutation - giving rise to an aggressive lymphoid malignancy.  When p53 mutations are present, chemotherapy often does not work well.  Just like every other cancer we have ever studied, p53 is a BAD THING to have mutated in DLBCL.  Conversely, when DLBCL develops spontaneously, it is often a curable cancer.  This plays out with regard to prognosis of the transformation.  In the former, survival averages about a year, whereas in the latter a good number of patients are cured.  Once again, I would point out that it is frustrating that we have no way to tell which one a patient has with testing that we would consider readily available.  In indolent lymphoma, it seems far more likely that the new DLBCL is clonally related and p53 mutations are as high as 80%!

Treatment often consists of R-CHOP chemotherapy regardless of which sort of RT you have.  While this is typically a well tolerated treatment, it is harder on the patient if they have already been exposed to a bunch of chemotherapy previously.  You only get to beat up the bone marrow so many times (chemo) before it starts telling you it can't accept more flogging.  It is not uncommon to run into dose delays, or reduced dosing, etc.  Add this to more resistant disease and you can probably figure why it is less effective.  Furthermore, a lot of patients with follicular lymphoma have been previously treated with the "H" in R-CHOP and you can only give so many doses of that drug before the heart starts to complain.  Since treatment is less effective, some patients will be treated with an "auto" stem cell transplant but a lot depends on how robust the patient is at that point and how well they responded to therapy.

In the future, I think this may be one situation where the "engineered T cells" could become an important therapy.  NOTCH antibodies have recently entered clinical trials and might be appealing.  We have used brentuximab vedotin in one clinical trial and been pleased with the results for some of our patients.  Hopefully these newer approaches will give a more favorable outlook to patients with RS/HT sometime soon.

Here is a video I did with Brian Koffman describing it all:  Richters Transformation