R-CHOP is one of the most commonly used regimens out there for lymphoma. I wanted to briefly write about what the different ingredients are - how they work - what patients experience - etc. R-CHOP actually is a cocktail of 5 different drugs. Each letter represents one of the drugs. Here goes...
R=Rituxan. Rituxan is an antibody. You make antibodies to fight colds, flu, e.coli, etc. Instead of a naturally occurring antibody, this one is "engineered" to bind to the outside of a lymphoma/leukemia cell and alert the immune system to go after it. See my other post "Building a better CD20 antibody." People can often have "infusion reactions" with the first dose (chills, shaking, shortness of breath, rash, etc). If you actually measure "b-cells" in the blood while administering the antibody, you can see them disappear from the blood during the infusion. As those b-cells go away they release little hormones that cause the symptoms. Often the symptoms do not recur with subsequent doses as the B cells are gone. Overall, most people tolerate the drug extremely well and are not even aware they are getting a very effective anti-cancer treatment. Those patients with side effects can often be managed by extra tylenol, benadryl, steroids, etc.
C=Cyclophosphamide. This is an old school chemotherapy that has been around for quite a few years. It binds to DNA in the nucleus of the cancer cell and prevents effective replication of the cells genome. Those cells that divide more rapidly are more sensitive to the treatment. Therefore cancer cells and normal bone marrow are most affected. It lowers healthy blood cells as well as bad ones. Fortunately the good guys recover more quickly. It can also cause nausea but our nausea medications are so good, that is rarely a problem. There can be bleeding in the bladder but I have given a ton of cytoxan and I've never seen it as a problem.
H=Hydroxyrubicin (which we call Adriamycin). Some call this drug "Big Red" as it has a pinkish / red color and can make your urine discolored for 24-48 hours. When you talk about R-CVP the only difference is that this drug is missing. Hydroxyrubicin "intercalates" or "slips between" the base pairs of DNA. DNA is a coiled structure. In order to copy it, you need to unzip the two strands. This creates a lot of "torque" or "strain" that needs to be released. It is like a hair dryer cord that gets turned around and bound up in knots. To relieve the strain there are enzymes called "topoisomerases" which nick the DNA so it can spin around and release the tension and then get re-attached. Adriamycin prevents that and results in "double stranded DNA breaks." These are harder for the cancer cell to repair than single stranded breaks. The side effect profile is notable for hair loss and low blood counts. The fear of adriamycin is that it can weaken the heart muscle. You can usually get away with 6-8 doses of the drug, but after that point the risk of heart damage really goes up. In uncommon situations, it can cause some permanent injury to your bone marrow stem cells and in extremely rare circumstances even cause a second cancer.... aargh!
O=Oncovin (which we call vincristine). Vincristine is what we call a "microtubule inhibitor." Other microtuble drugs include vinblastine, taxol, taxotere, eribulin, etc. Microtubules are like the cellular "skeleton." It helps hold all the pieces of the cell where they want to be. When cells need to separate their chromosomes into daughter cells they use microtubules to help pull them apart to the opposite sides of the cell. They chew up one end and add pieces to the other which creates a sense of "movement." Vincristine freezes that process and cells don't like that. If they stay frozen too long they trigger cell death. It causes some numbness in finger tips and toes and possibly constipation.
P=Prednisone (yes the good old fashioned steroid). You are probably familiar with prednisone for things like rashes or arthritis etc. We use it to lower the immune system. Lymphoma is simply a cancer of the immune system. I have had occasion to "cool off" a patient while trying to collect additional tests. I can sometimes give them a slug of steroids while I am waiting and see their lymph nodes shrink.
Sometimes when we tell patients about potential side effects, I get the impression the cure sounds a lot worse than the disease. Doc's probably practice defensive medicine more than we realize. If we tell you everything that can go wrong, then we feel we are less likely to get sued if something bad happens. Unfortunately you see some patients go through the side effect list and freak out - understandably.
I often say that chemotherapy is a lot like a "stereo dial." We can turn it up to 10 (unbearable) or down to 1 (was that chemo you just gave me?). R-CHOP is somewhere between a 4-6. Nausea isn't often much of an issue. Fatigue is considerable. Many people can work while getting chemo, but some folks are better off taking the time off of work (if they can). Somehow the hair loss makes you feel like a cancer patient. After you have had one dose of chemo (a cycle is every three weeks you get one dose) I tell patients that they have seen 80% of the side effects. Yes there is some degree to which things are cumulative, but for the most part you have seen the side effects.
Hopefully that helps. In another post, I will talk about how I approach DLBCL and when I choose alternatives.
Friday, August 31, 2012
Friday, August 24, 2012
Making sense of all the different lymphomas
I am not sure if there is a single person on the planet who can tell me exactly how many different types of lymphoma exist. I live and breath lymphoma and I can't. Perhaps Elaine Jaffe M.D. could since she is the pathologist in charge of the world health organization committee responsible for lymphoma nomenclature, but I would wager that she might pause before giving you an answer. I think the number is somewhere between 70-100 different types - but it sort of depends how far into the weeds you want to get with your counting.
Take the most common form of NHL known as Diffuse Large B Cell Lymphoma (aka DLBCL) and you could have such variants as primary CNS (brain) DLBCL, primary mediastinal (between the lungs) DLBCL, intravascular DLBCL, gastric (stomach) DLBCL, activated B cell DLBCL, germinal center DLBCL, DLBCL of leg, etc. Remarkably each of these entities has a fairly significantly different biology / prognosis and they warrant unique classification. It is not merely the part of the body where they show up, they are actually different diseases.
If you even venture to take a single entity like "activated B-cell DLBCL" there may be key molecular differences that could split this disease into multiple subtypes depending on whether there is a CD79b mutation versus a CARD11 mutation or a MYD88 mutation which could be the key difference in responses to drugs like ibrutinib.
This bewildering complexity is one of the reasons I most enjoy being involved in lymphoma / CLL. Even amongst seasoned oncologists, you can often look like the brightest person in the room when you talk about subtle patient management differences in uncommon variants.
Unfortunately this bewildering complexity can also be overwhelming for a patient who is told by their surgeon that they have lymphoma and need to see an oncologist. The delay between the diagnosis and the consult is often filled with an internet search that feels like a trip through Alice in Wonderland.
When I meet a patient with lymphoma I always explain how these different diseases fall into a few limited categories and how those categories are unique. Hopefully you will find this useful.
The first "branch point" is the difference between Hodgkin's Lymphoma and Non-Hodgkin's Lymphoma. Patients always ask, "Which is better?" Of course there are a lot of variables that determine the answer to that so I usually say, "stick with me for a minute... we will get there."
Among the patients with Non-Hodgkin's Lymphoma (which outnumber Hodgkin's by about 5:1), my next branch point is T cell versus B cell. In normal physiology, B cells are responsible for making antibodies and T cells are either the "brains" of the immune system or the "cellular assassins" that go around looking for a rogue cell to annihilate. B cell lymphoma outnumbers T cell by about 9:1 so when most people talk about lymphoma they are talking about one of the subtypes of B cell Non-Hodgkin's Lymphoma.
So if we focus for now on the different types of B cell Non Hodgkin's Lymphoma the next branch point splits into three categories based more on clinical behavior. I call them "slow, medium, and fast." In official parlance they are "indolent, aggressive, highly aggressive." Within each category, there is one form of NHL that is the most common and it becomes the "prototype" of the category - for better or for worse. Slow = Follicular Lymphoma (all my FL posts can be linked here), Medium = Diffuse Large B Cell Lymphoma, Fast = Burkitt's. There are a bunch of other slow lymphomas such as Marginal Zone Lymphoma, MALT lymphoma, Small Lymphocytic Lymphoma, etc. There can be considerable differences in management between these, but there is still value in lumping them together as a category. For example, if you learn the basic features of follicular lymphoma (slow prototype), you have a good place to start then you can learn how your specific disease is different or unique.
It is important to recognize however that even a singular entity like Follicular Lymphoma can range from a disease that never needs treatment to a disease that is immediately life threatening. There are other diseases like Mantle Cell Lymphoma that don't fit nicely into a singular category and can take on properties of both "slow" or "medium" lymphomas.
Back in medical school, they always taught us how important a patients history is. In oncology I find this to be most true when I evaluate a new patient with lymphoma. If that lymph node sprung up in 6 days, that is very different than a lymph node that sprung up in 6 months, or even the occasional patient who comes in and says that the lymph node has been there for 6 years. Yes there are categories, but you cannot take a variable that has a continuous distribution and neatly put it into discrete categories!
With all those limitations, I still want to make an effort to say how the categories are different.
If you go from slow to medium to fast I tell patients that the "curability" increases and the "managability" decreases. You are probably not going to cure a slow lymphoma, but for quite a few patients you can manage it for many years. In the Rummel presentation at ASCO 2012, the median (half of patients do better, half do worse) progression free survival after bendamustine / rituxan was almost 7 years - even without rituxan maintenance. That doesn't even take into account survival based on second line regimens and beyond. Slow lymphoma is often very manageable. I have quite a few patients out many years who have NEVER REQUIRED ANY TREATMENT! This is the group where you often hear "watch and wait." I understand that patients want to be "cured." Sadly, with our current treatment regimens I don't think patients with slow lymphoma often get there (ok ignore the minority of limited stage slow lymphoma that gets radiated doesn't have their disease come back elsewhere). I often use the analogy of having a "quiver of arrows." You use the arrows as you need them and count on research to add arrows to your quiver as you go. With survival that is often over a decade, that does not feel to me like a false hope.
If you go the other direction from fast to medium to slow, "managibility" increases. On the highly aggressive end of the spectrum are Burkitt's lymphoma or Lymphoblastic Lymphoma. Here you are aiming for cure but if you don't get there you are in trouble because it is unlikely to be very manageable. Often you don't meet a patient with Burkitt's in the clinic. They often get hospitalized because something has gone terribly wrong and they are hospitalized before the diagnosis is made. Other times, you get a call from a primary care physician who has made the diagnosis and is requesting a consult. I tell them to send the patient straight to the hospital, I will meet them there! From the outset, you are fighting a high stakes battle. With these diseases you often have one chance to get it right. If things go wrong, your chance for cure gets a lot more difficult.
DLBCL is somewhere in between. It is the most common form of NHL (about half of all cases). The disease is quick, but often you have enough time to dot all your "i's" and cross all your "t's" before launching into chemotherapy which probably cures a little more than half of all patients. "Watch and Wait" does not apply to this group, but you often have time to get your PET scan, bone marrow biopsy, echocardiogram, etc. as an outpatient. For those patients not cured, it can often be in remission for years. Even if it comes back, you often have a back-up plan that can cure them the second time around (albeit with a "auto stem cell transplant"). Not surprisingly though there is a broad range of behavior among cases of DLBCL. In some cases it acts like a bad case of Follicular and at other times it can give Burkitt's a run for the money.
I've largely ignored T cell NHL to this point. Though there are "highly aggressive" T cell lymphomas they are fairly rare. Most of the systemic T cell disease fall into the same "intermediate / medium" category as DLBCL. There are a variety of skin T cell lymphomas that are in the "slow" category. Start talking T cell lymphoma variants and you often get a glazed look from oncologists who don't think about NHL all that often.
To summarize: we often lump NHL into one of three categories. Though there can be a significant range within an individual category, our clinical approach to each category is unique and the expectations of patient outcome varies by category. The prototypes (most common per category) are follicular (30%), DLBCL (40-50%), Burkitt's (5%). The remainder of cases make up the bewildering complexity of NHL and explain why my job will likely never be outsourced....
Take the most common form of NHL known as Diffuse Large B Cell Lymphoma (aka DLBCL) and you could have such variants as primary CNS (brain) DLBCL, primary mediastinal (between the lungs) DLBCL, intravascular DLBCL, gastric (stomach) DLBCL, activated B cell DLBCL, germinal center DLBCL, DLBCL of leg, etc. Remarkably each of these entities has a fairly significantly different biology / prognosis and they warrant unique classification. It is not merely the part of the body where they show up, they are actually different diseases.
If you even venture to take a single entity like "activated B-cell DLBCL" there may be key molecular differences that could split this disease into multiple subtypes depending on whether there is a CD79b mutation versus a CARD11 mutation or a MYD88 mutation which could be the key difference in responses to drugs like ibrutinib.
This bewildering complexity is one of the reasons I most enjoy being involved in lymphoma / CLL. Even amongst seasoned oncologists, you can often look like the brightest person in the room when you talk about subtle patient management differences in uncommon variants.
Unfortunately this bewildering complexity can also be overwhelming for a patient who is told by their surgeon that they have lymphoma and need to see an oncologist. The delay between the diagnosis and the consult is often filled with an internet search that feels like a trip through Alice in Wonderland.
When I meet a patient with lymphoma I always explain how these different diseases fall into a few limited categories and how those categories are unique. Hopefully you will find this useful.
The first "branch point" is the difference between Hodgkin's Lymphoma and Non-Hodgkin's Lymphoma. Patients always ask, "Which is better?" Of course there are a lot of variables that determine the answer to that so I usually say, "stick with me for a minute... we will get there."
Among the patients with Non-Hodgkin's Lymphoma (which outnumber Hodgkin's by about 5:1), my next branch point is T cell versus B cell. In normal physiology, B cells are responsible for making antibodies and T cells are either the "brains" of the immune system or the "cellular assassins" that go around looking for a rogue cell to annihilate. B cell lymphoma outnumbers T cell by about 9:1 so when most people talk about lymphoma they are talking about one of the subtypes of B cell Non-Hodgkin's Lymphoma.
So if we focus for now on the different types of B cell Non Hodgkin's Lymphoma the next branch point splits into three categories based more on clinical behavior. I call them "slow, medium, and fast." In official parlance they are "indolent, aggressive, highly aggressive." Within each category, there is one form of NHL that is the most common and it becomes the "prototype" of the category - for better or for worse. Slow = Follicular Lymphoma (all my FL posts can be linked here), Medium = Diffuse Large B Cell Lymphoma, Fast = Burkitt's. There are a bunch of other slow lymphomas such as Marginal Zone Lymphoma, MALT lymphoma, Small Lymphocytic Lymphoma, etc. There can be considerable differences in management between these, but there is still value in lumping them together as a category. For example, if you learn the basic features of follicular lymphoma (slow prototype), you have a good place to start then you can learn how your specific disease is different or unique.
It is important to recognize however that even a singular entity like Follicular Lymphoma can range from a disease that never needs treatment to a disease that is immediately life threatening. There are other diseases like Mantle Cell Lymphoma that don't fit nicely into a singular category and can take on properties of both "slow" or "medium" lymphomas.
Back in medical school, they always taught us how important a patients history is. In oncology I find this to be most true when I evaluate a new patient with lymphoma. If that lymph node sprung up in 6 days, that is very different than a lymph node that sprung up in 6 months, or even the occasional patient who comes in and says that the lymph node has been there for 6 years. Yes there are categories, but you cannot take a variable that has a continuous distribution and neatly put it into discrete categories!
With all those limitations, I still want to make an effort to say how the categories are different.
If you go from slow to medium to fast I tell patients that the "curability" increases and the "managability" decreases. You are probably not going to cure a slow lymphoma, but for quite a few patients you can manage it for many years. In the Rummel presentation at ASCO 2012, the median (half of patients do better, half do worse) progression free survival after bendamustine / rituxan was almost 7 years - even without rituxan maintenance. That doesn't even take into account survival based on second line regimens and beyond. Slow lymphoma is often very manageable. I have quite a few patients out many years who have NEVER REQUIRED ANY TREATMENT! This is the group where you often hear "watch and wait." I understand that patients want to be "cured." Sadly, with our current treatment regimens I don't think patients with slow lymphoma often get there (ok ignore the minority of limited stage slow lymphoma that gets radiated doesn't have their disease come back elsewhere). I often use the analogy of having a "quiver of arrows." You use the arrows as you need them and count on research to add arrows to your quiver as you go. With survival that is often over a decade, that does not feel to me like a false hope.
If you go the other direction from fast to medium to slow, "managibility" increases. On the highly aggressive end of the spectrum are Burkitt's lymphoma or Lymphoblastic Lymphoma. Here you are aiming for cure but if you don't get there you are in trouble because it is unlikely to be very manageable. Often you don't meet a patient with Burkitt's in the clinic. They often get hospitalized because something has gone terribly wrong and they are hospitalized before the diagnosis is made. Other times, you get a call from a primary care physician who has made the diagnosis and is requesting a consult. I tell them to send the patient straight to the hospital, I will meet them there! From the outset, you are fighting a high stakes battle. With these diseases you often have one chance to get it right. If things go wrong, your chance for cure gets a lot more difficult.
DLBCL is somewhere in between. It is the most common form of NHL (about half of all cases). The disease is quick, but often you have enough time to dot all your "i's" and cross all your "t's" before launching into chemotherapy which probably cures a little more than half of all patients. "Watch and Wait" does not apply to this group, but you often have time to get your PET scan, bone marrow biopsy, echocardiogram, etc. as an outpatient. For those patients not cured, it can often be in remission for years. Even if it comes back, you often have a back-up plan that can cure them the second time around (albeit with a "auto stem cell transplant"). Not surprisingly though there is a broad range of behavior among cases of DLBCL. In some cases it acts like a bad case of Follicular and at other times it can give Burkitt's a run for the money.
I've largely ignored T cell NHL to this point. Though there are "highly aggressive" T cell lymphomas they are fairly rare. Most of the systemic T cell disease fall into the same "intermediate / medium" category as DLBCL. There are a variety of skin T cell lymphomas that are in the "slow" category. Start talking T cell lymphoma variants and you often get a glazed look from oncologists who don't think about NHL all that often.
To summarize: we often lump NHL into one of three categories. Though there can be a significant range within an individual category, our clinical approach to each category is unique and the expectations of patient outcome varies by category. The prototypes (most common per category) are follicular (30%), DLBCL (40-50%), Burkitt's (5%). The remainder of cases make up the bewildering complexity of NHL and explain why my job will likely never be outsourced....
Wednesday, August 22, 2012
Campath is going away?
I was quite surprised today to hear that Campath is being withdrawn from the oncology marketplace. This will affect some patients with CLL and some patients with cutaneous lymphoma
Campath is a "bit player" in the management of CLL/NHL but for docs comfortable using it - it can be quite handy. It is one of the few drugs that clearly works in the group of patients with 17p deletion and can be effective when other drugs have failed provided the patient does not have bulky lymph nodes (Campath is good at marrow and circulating cells but doesn't often resolve big nodes).
We have always known that Campath profoundly lowers the immune system. Unfortunately, some of the initial studies were done in patients who were already heavily pre-treated. It probably made the drug look worse than it really is to study it in patients whose immune system was already beaten up by multiple rounds of fludarabine. None the less, you have to be on guard for all sorts of unusual infections such as CMV, PCP, shingles, etc.
Viagara was initially supposed to be a drug to treat hypertension and when men had an unusual "side effect" while taking the drug, it quickly got repurposed and became an extremely valuable asset for Pfizer. Similarly, there are reasons other than CLL where you may wish to lower the immune system. Multiple Sclerosis is just such a reason.
As is the case with lots of drugs - dose matters! They are hoping to show that lower campath doses can treat MS and avoid the infectious complications. If that works, it will probably be a much more valuable drug for its owners. Campath will be "re-branded" so I am not sure if it will still be available for patients with CLL.
To me, this is potentially bad news because although it was never much more than 5% of CLL patients ever treated with the drug - for that 5% it was quite good. It sounds like there may be patient access programs for patients with CLL - hopefully those materialize. Rumor has it that it may even be given away for free to patients with CLL - obviously that would be a good thing - but I will believe it when I see it. Furthermore, additional barriers may just make a hard to use drug even harder to get.
We are lucky that the CAL-101/GS-1101, ibrutinib drugs appear to have activity in 17p deleted cases - these drugs are coming just in time to help.
Campath is a "bit player" in the management of CLL/NHL but for docs comfortable using it - it can be quite handy. It is one of the few drugs that clearly works in the group of patients with 17p deletion and can be effective when other drugs have failed provided the patient does not have bulky lymph nodes (Campath is good at marrow and circulating cells but doesn't often resolve big nodes).
We have always known that Campath profoundly lowers the immune system. Unfortunately, some of the initial studies were done in patients who were already heavily pre-treated. It probably made the drug look worse than it really is to study it in patients whose immune system was already beaten up by multiple rounds of fludarabine. None the less, you have to be on guard for all sorts of unusual infections such as CMV, PCP, shingles, etc.
Viagara was initially supposed to be a drug to treat hypertension and when men had an unusual "side effect" while taking the drug, it quickly got repurposed and became an extremely valuable asset for Pfizer. Similarly, there are reasons other than CLL where you may wish to lower the immune system. Multiple Sclerosis is just such a reason.
As is the case with lots of drugs - dose matters! They are hoping to show that lower campath doses can treat MS and avoid the infectious complications. If that works, it will probably be a much more valuable drug for its owners. Campath will be "re-branded" so I am not sure if it will still be available for patients with CLL.
To me, this is potentially bad news because although it was never much more than 5% of CLL patients ever treated with the drug - for that 5% it was quite good. It sounds like there may be patient access programs for patients with CLL - hopefully those materialize. Rumor has it that it may even be given away for free to patients with CLL - obviously that would be a good thing - but I will believe it when I see it. Furthermore, additional barriers may just make a hard to use drug even harder to get.
We are lucky that the CAL-101/GS-1101, ibrutinib drugs appear to have activity in 17p deleted cases - these drugs are coming just in time to help.
Saturday, August 18, 2012
Building a better CD20 antibody?
Rituxan was an unbelievable breakthrough drug in the treatment of all sorts of B cell lymphoid cancers (CLL / NHL) as well as a lot of autoimmune conditions to boot. At the time it was put into the patients, the idea of using therapeutic antibodies to treat cancer had not really been done. Skeptics abounded. As soon as it was tried, it was immediately obvious that there had been a breakthrough. One theme I continually revist is that the immune system can be a very powerful cancer fighter when properly directed. Rituxan was one of the first ways to focus this aspect of the immune system to fight off lymphoma.
You make antibodies to fight off the flu, e. coli, etc. Instead of the naturally occuring variety of antibodies, rituxan is a synthesized antibody designed to fight off B-Cells (ie. the cells that give rise to lymphoma and CLL). Rituxan binds onto the outside of the cancer cell (on a marker known as CD20) and has about 4-5 ways to kill the cell.
15 years later, hundreds of thousands of patients have been treated. As a single agent nearly 70% of patients with untreated follicular lymphoma will respond and in many cases the responses can be very durable. In most situations where giving chemotherapy for CLL/NHL is good, giving it together with rituxan is better. Not bad for a drug that has few of the side effects of traditional chemotherapy!
Not long after Rituxan demonstrated its value, lots of science went toward understanding why rituxan does or doesn't work. With the accumulated knowledge, different companies have sought to make a better version of the drug.
Ofatumumab was the next CD20 antibody to be approved. The main improvements in ofatumumab was more sturdy binding to CD20, binding at a different location on CD20, and a better job "fixing complement." Complement is a set of "executioner proteins" that get recruited from the circulation when there is enough antibody bound to the outside of a cell. "Fixing complement" just means it does a better job attracting those proteins to bind to the outside of the cell and punch holes through the membrane.
GA-101 is an ant-CD20 antibody currently in clinical trials in both CLL/NHL. GA-101 (aka obinutuzumab) does several things new. It binds to CD20 in a different orientation. Imagine an antibody like the letter "Y" The top of the "Y" is the little "grabber" part that identifies its target while the bottom part of the "Y" sticks out and alerts the immune system. GA-101 binds more vertically which sticks the action end out there better for the rest of the immune system to see. Rituxan binds more flat to the cell and the action end sort of gets burried on the surface of the cell. GA-101 also has a more flexible "hinge region" (where the "Y" splits) which is supposed to help bring together the different proteins better. GA-101 also "engineers out" the complement activity and instead "engineers in" better ADCC. ADCC (antibody dependent cellular cytotoxicity) is the term that describes when a "killer cell" attackes a target cell because it is coated with antibody. Finally there is a somewhat poorly characterized process by which antibodies can cause direct cell killing. This is supposedly enhanced by antibodies known as "Type 2" antibodies (not really worth explaining this one, except that Rituxan / Ofatumumab are Type 1)
So far, there are not a lot of studies that clinically compare the different anti CD20 antibodies head to head. Ofatumumab initially tried to get an FDA approval in follicular lymphoma patients considered refractory to Rituximab. Results didn't really impress all that much. Based upon FDA approval strategies, it also tried to gain approval in CLL patients were refractory to fludarabine AND campath (or bulky fludarabine refractory where campath doesn't work well). This worked better and led to the approval of ofatumumab in CLL. The drug has not yet gained a high utilization in CLL.
GA-101 is owned by Genentech and with Rituxan likely to go off patent in the future, they are working hard to make GA-101 the standard-bearer. At ASH 2011 there were a lot of clinical papers with GA-101 but the most notable was a head to head comparison of GA-101 to Rituxan in patients with relapsed follicular lymphoma. GA-101 had twice the response rate though the duration of response was not a lot different. Hard to say if that is a glass half full or half empty - trying to extrapolate to different NHL / CLL situations it may depend on the context. Increasing a response rate in DLBCL might be very important whereas PFS prolongation may be more important in follicular lymphoma. Furthermore, we should probably consider that the activity of one antibody may be better in CLL while a different set of characteristics may make a different antibody better in follicular lymphoma.
Perhaps the cleanest way to figure if a drug is active is in the previously untreated patients. Ofatumumab was studied in this setting and reported at ASH. GA-101 has been evaluated in front line CLL and I found this news story anecdote to be pretty exciting though you must be very careful with anecdotes and no formal reports are available. Rituximab was evaluated in relapsed patients way back in the early 2000s with pretty modest single agent activity, though improved with higher dosing of the drug or greater frequency (another way of increasing the dose).
You can bet you will hear a lot about "better versions of rituxan." Look for head to head comparisons or results that are substantially different from prior series. Rituxan was definitely a pioneer, but I fully expect we will be able to build a better treatment. Here is a review published in ASH that goes over a lot of the information above and describes some of the other strategies to improve upon rituxan.
Thursday, August 16, 2012
How do ibrutinib and idelalisib work?
I've had a lot of traffic on my site from patients looking for information on ibrutinib and CAL-101. I thought that the paper published by Hassan Jumaa this week helped shed more light on the unique activity in CLL:
Chronic lymphocytic leukaemia is driven by antigen-independent cell-autonomous signalling
The substance of the article is that the B-Cell Receptor (aka antibody or BCR) from patients with CLL continuously triggers cell growth because it reacts with itself. The BCR is the one of the main determinants of B cell fate. It has long been suspected that the CLL cells may encode a BCR which recognizes either something inside of a patients body or perhaps even is triggered by an external stimulus that is common such as Tom Kipps earlier paper indicating that it might be part of the CMV virus that is recognized.
Recombinant antibodies encoded by IGHV1-69 react with pUL32, a phosphoprotein of cytomegalovirus and B-cell superantigen
In either case, constant signaling through the BCR gives an unending "growth signal" to the cancer cells. It adds to the argument that the BCR signaling pathway therefore represents a key "node" in B cell survival - ie. it is one of the main "pro-growth" factors in CLL. It follows therefore that drugs that interrupt this pathway would interfere with the growth / survival signals given by the receptor.
Since the first three enzymes in the pathway are Syk, BTK, and PI3K it is not a surprise that drugs affecting these three enzymes act in similar manner.
Double click on the image if you actually want to be able to see it.
Fostamatinib (syk) led the way and showed that the target was of value but the drug has had a difficult time gaining traction as the main development focus of the drug was for rheumatoid arthritis.
Richard Miller (CEO of pharmacyclics at the time and occasional attending physician at Stanford) had the opportunity to see some of our patients respond to fostamatinib. His company was developing a drug PCI-32765 (only recently named ibrutinib) which in hibited the next enzyme in the pathway - BTK. He got it polished up for a clinical trial which we developed over several weeks. Ibrutinib is a remarkable drug in CLL and much has been written about it. Much more will be written about it.
About the same time a small Seattle Bio-pharmaceutical company had created a drug that inhibits PI3K - the next enzyme in the pathway. PI3K is a little different in that it integrates important signals from a variety of cell surface receptors. Those signals that tend to make a cell survive turn on PI3K and those that signal cell death turn off PI3K. PI3K is actually a large family of proteins but the "delta-isoform" seems to be specific for B cells. Their drug initially called CAL-101 (after one of the early founders of the company turned onto highway 101 in Palo Alto and decided that would be the name of the drug) now renamed GS-1101 (after acquisition by Gilead) shuts off PI3K and has considerable reported activity in CLL and NHL.
The point is the following: CLL cells proliferate in response to signaling through the B cell receptor. Drugs that target the pathway that transmits the signal inside of the cell turn off that proliferation and make the cancer cells less able to survive. The new finding in Hasaan's paper is that the source of the signal may be that the B cell receptor reacts with itself instead of the invading micro-organism it was supposed to attack.
Chronic lymphocytic leukaemia is driven by antigen-independent cell-autonomous signalling
The substance of the article is that the B-Cell Receptor (aka antibody or BCR) from patients with CLL continuously triggers cell growth because it reacts with itself. The BCR is the one of the main determinants of B cell fate. It has long been suspected that the CLL cells may encode a BCR which recognizes either something inside of a patients body or perhaps even is triggered by an external stimulus that is common such as Tom Kipps earlier paper indicating that it might be part of the CMV virus that is recognized.
Recombinant antibodies encoded by IGHV1-69 react with pUL32, a phosphoprotein of cytomegalovirus and B-cell superantigen
In either case, constant signaling through the BCR gives an unending "growth signal" to the cancer cells. It adds to the argument that the BCR signaling pathway therefore represents a key "node" in B cell survival - ie. it is one of the main "pro-growth" factors in CLL. It follows therefore that drugs that interrupt this pathway would interfere with the growth / survival signals given by the receptor.
Double click on the image if you actually want to be able to see it.
Fostamatinib (syk) led the way and showed that the target was of value but the drug has had a difficult time gaining traction as the main development focus of the drug was for rheumatoid arthritis.
Richard Miller (CEO of pharmacyclics at the time and occasional attending physician at Stanford) had the opportunity to see some of our patients respond to fostamatinib. His company was developing a drug PCI-32765 (only recently named ibrutinib) which in hibited the next enzyme in the pathway - BTK. He got it polished up for a clinical trial which we developed over several weeks. Ibrutinib is a remarkable drug in CLL and much has been written about it. Much more will be written about it.
About the same time a small Seattle Bio-pharmaceutical company had created a drug that inhibits PI3K - the next enzyme in the pathway. PI3K is a little different in that it integrates important signals from a variety of cell surface receptors. Those signals that tend to make a cell survive turn on PI3K and those that signal cell death turn off PI3K. PI3K is actually a large family of proteins but the "delta-isoform" seems to be specific for B cells. Their drug initially called CAL-101 (after one of the early founders of the company turned onto highway 101 in Palo Alto and decided that would be the name of the drug) now renamed GS-1101 (after acquisition by Gilead) shuts off PI3K and has considerable reported activity in CLL and NHL.
The point is the following: CLL cells proliferate in response to signaling through the B cell receptor. Drugs that target the pathway that transmits the signal inside of the cell turn off that proliferation and make the cancer cells less able to survive. The new finding in Hasaan's paper is that the source of the signal may be that the B cell receptor reacts with itself instead of the invading micro-organism it was supposed to attack.
Wednesday, August 8, 2012
Trisomy 12 in CLL / Notch Mutations
Another common FISH abnormality in CLL is trisomy 12 (literally "three body") in which there are three copies of chromosome 12 whereas the normal allocation is just two copies. There has been a lot of recent excitement about additional abnormalities found in trisomy 12 CLL but I wager there is a lot more to be found. Even though chromosome 12 is a "mid-size" chromosome, there are literally thousands of genes on the chromosome so I expect we will find more information in the near future.
Before we had FISH technology and the only way to study chromosomes was "metaphase cytogenetics" where you literally grew up the cells in a dish, took out the chromosomes, froze them just before the cell was going to divide (ie. metaphase) and visually inspected them, trisomy 12 was one of the most common abnormalities found. With FISH, we've become a lot better at finding chromosome 13q changes so it has fallen from the top spot.
Trisomy 12 CLL has a distinct personality among cases of CLL. When you look at the cells under the microscope, they have a lot more "atypical" features. In an individual patient, there may be a higher frequency of "prolymphocytes" which look a little more aggressive.
There are several key clinical features of trisomy 12. One feature I think is intersting though I have not seen a lot of literature about it is that it has a considerabally higher amount of CD20 on the surface of the cell. CD20 is the target of rituximab / ofatumumab / GA-101 and a variety of other research antibodies. One of the reasons we don't think rituxan is as effective as a single agent in CLL compared to most of the lymphomas is that it has a lot lower CD20 on the surface of the cell. The following paper actually suggests a higher response rate to single agent rituximab in cases with trisomy 12 (actually they also gave gm-csf but the rituxan did most the heavy lifting).
Chronic lymphocytic leukaemia CD20 expression is dependent on the genetic subtype: a study of quantitative flow cytometry and fluorescent in-situ hybridization in 510 patients
This CD20 differential may become important as combination biologic regimens begin to emerge. I look forward to anti-CD20 in combination with BCR signaling inhibitors or BCL-2 inhibitors. It would be very interesting to me if those patients with trisomy 12 did better under those conditions. That is purely speculative on my part for now, but something I would like to investigate.
In the landmark Dohner paper in New England Journal of Medicine that identified impact of FISH on survival in CLL, trisomy 12 was considered an "intermediate" prognostic feature.
Genomic Aberrations and Survival in Chronic Lymphocytic Leukemia
It has recently been discovered however that a considerable fraction of trisomy 12 cases (somewhere between 1/3 to 1/2 may have a mutation in a key protein called NOTCH. A mutation in NOTCH is a really bad thing - just about as bad as having a mutation or deletion of p53 / 17p, or another bad actor Myc. We don't currently have convenient tests for testing for NOTCH unfortunately. I wager though that once we do, trisomy 12 will either be a good prognostic factor (in setting of normal NOTCH), or bad factor (when NOTCH is mutated)
NOTCH1 mutations in +12 chronic lymphocytic leukemia (CLL) confer an unfavorable prognosis, induce a distinctive transcriptional profiling and refine the intermediate prognosis of +12 CLL
NOTCH is a protein at the surface of the CLL cell. It waves its receptor out to other cells and when it comes into contact with the right signal it causes the part of the protein on the inside of the cell (conveniently called the intracellular domain aka ICD) to be released from the membrane to go into the nucleus of the cell where it turns on a lot of different genes related to cellular proliferation. It goes about turning on genes until another protein called FBXW7 comes around and attaches a tag on NOTCH that targets it for destruction. Interestingly FBXW7 has also recently been identified as a gene periodically mutated in CLL which is a well developed theme - there are lots of ways to activate the same pathway.
Activation of the NOTCH1 pathway in chronic lymphocytic leukemia
There are a handful of drugs out there that target the NOTCH pathway. One way to target the pathway is to inhibit the enzyme that cuts it free from the cell membrane. These are so called "gamma secretase inhibitors." Unfortunately, this enzyme is important in the bowel and diarrhea has been a big problem. There are other ways to go after the NOTCH protein and the hope is that some of them may not have the same GI issues.
I bet in time, we find a number of other abnormalities associated with trisomy 12. Whether there are other miRNA's or other features. In the quest toward personalized medicine this is the level of detail that will be required.
Sunday, August 5, 2012
Clonal Evolution in CLL
This article came out about the same time as one from John Byrd's lab. Both show the same concept and I think it is important. It highlights the evolutionary capability of CLL over time. The disease is not the same disease at initial diagnosis and relapse.
You often hear of patients say that when their disease was diagnosed they were 13q but when it came back it may be both 13q and 17p. Frankly, FISH is only a small portion of the story. There are a lot of point mutation in key genes such as SF3B1, BIRC3, NOTCH, etc. For the time being, we don't have an easy way to even test for these abnormalities let alone monitor over time.
I think it is a lot like when you took antibiotics for your sore throat as a child. You were told to take all your medicine so that the bacteria didn't get resistant. I think of CLL as the germ that keeps coming back a little smarter than before.
When I did lab work, I was amazed how many CLL cells you could take out of a single tube of a patients blood. If a patient had a high white blood cell count, you could literally get tens of millions of cancer cells in a single 10mL tube of blood. In many cases we know that these cancer cells are "genomically unstable" (especially 11q, 17p) which means they are all free to "experiment" with new mutations. Throw in all the blood, the marrow, the spleen and you have many billions of cells.
Now treat that patient.
Let's say you have effective therapy and get rid of 99.99% of the cells (ie. only one out of 10,000 survive - perhaps we should write another post about minimal residual disease), you may have millions of cells left over that survived because they had the right genetic changes to survive the treatment. Those are the ones that may then have a competitive advantage and establish themselves as the dominant clone at relapse.
This potentially has broad implications on things such as initial therapy selection, treatment intensity, elimination of minimal residual disease, etc. I am not sure we have the tools to deploy on broad enough scale just yet to answer these questions but with genomic sequencing coming online so fast, perhaps that will change soon.
You often hear of patients say that when their disease was diagnosed they were 13q but when it came back it may be both 13q and 17p. Frankly, FISH is only a small portion of the story. There are a lot of point mutation in key genes such as SF3B1, BIRC3, NOTCH, etc. For the time being, we don't have an easy way to even test for these abnormalities let alone monitor over time.
I think it is a lot like when you took antibiotics for your sore throat as a child. You were told to take all your medicine so that the bacteria didn't get resistant. I think of CLL as the germ that keeps coming back a little smarter than before.
When I did lab work, I was amazed how many CLL cells you could take out of a single tube of a patients blood. If a patient had a high white blood cell count, you could literally get tens of millions of cancer cells in a single 10mL tube of blood. In many cases we know that these cancer cells are "genomically unstable" (especially 11q, 17p) which means they are all free to "experiment" with new mutations. Throw in all the blood, the marrow, the spleen and you have many billions of cells.
Now treat that patient.
Let's say you have effective therapy and get rid of 99.99% of the cells (ie. only one out of 10,000 survive - perhaps we should write another post about minimal residual disease), you may have millions of cells left over that survived because they had the right genetic changes to survive the treatment. Those are the ones that may then have a competitive advantage and establish themselves as the dominant clone at relapse.
This potentially has broad implications on things such as initial therapy selection, treatment intensity, elimination of minimal residual disease, etc. I am not sure we have the tools to deploy on broad enough scale just yet to answer these questions but with genomic sequencing coming online so fast, perhaps that will change soon.
Friday, August 3, 2012
How I approach follicular lymphoma (part 2) - patients with higher risk disease
In the last follicular lymphoma post, we talked about what to do with our patients with lower risk or less extensive disease. I wanted to talk about patients with more extensive disease or higher risk features.
If we go back to the lymphocare study as a starting point, back in 2004-2009 timeframe the main rituxan containing chemotherapy regimen choices were R-CHOP > R-CVP > R-Fludarabine.
With the genius of drug regimen naming, I should point out that R-CVP and R-CHOP are essentially the same drugs (vincristine is also called oncovin) with the latter having the addition of hydroxydoxorubicin which we call adriamycin - aargh. R-CVP is mild but the addition of adriamycin is a pretty big deal for most patients. It causes you to lose your hair, it can possibly weaken your heart muscle, and it can increase the likelihood of long term damage to your bone marrow. These regimens were built out of the early chemotherapy days when "combination chemo" was all the rage. The idea was that if one drug was good, four must be better.
To my knowledge there was never a CVP vs CHOP study sans rituxan (though if an reader knows of one, I will be happy to update my post). In the pre-rituximab era CVP became the regimen you used when you had less extensive disease or you were worried about someone being too frail for CHOP. CHOP was the "big guns" and used quite extensively to maximize remission duration.
CVP and CHOP both got a good "facelift" with the addition of rituxan which joined the picture in the very late 1990's. Rituxan was the perfect drug to add to these combinations since it had essentially no toxicity on the bone marrow, was very effective on it's own, and seemed to make chemo work better. There were several studies in which the design was chemo +/- rituxan - I highlight two:
This year at ASCO was the first time I have seen a direct comparison between R-CVP and R-CHOP (ignore the fludarabine-mitoxantrone combo because it is infrequently used in the US). I don't think this study could have been done in the US because a study in which you are randomized to adriamycin or not is a tough sell to patients - thank goodness for the Italians!
This turned a lot of heads when it was initially presented at ASH in 2009 because it showed that BR was at least equivalent to R-CHOP. Since BR is considerably easier on the patient (no hair loss, no cardiac risk, perhaps less heme side effects) - it would be the equivalent of a mid weight boxer taking a heavy weight boxer into the late rounds for a tie match.
Perhaps more shocking however was the updated ASCO presentation this last June when it captured the plenary session for demonstrating a convincingly superior "progression free survival" (time that patient is alive without recurrence of lymphoma). By a wide margin (average 3 years vs 6 years) BR kept patients disease away for longer with less side effects.
I have heard an interesting criticism of the study that the R-CHOP arm performed far worse in this study than virtually all other studies in FL. Perhaps the magnitude of the difference is smaller than what is reported, but in a randomized study, you would expect BR to end up underperforming as well so I am not sure how to interpret that observation.
So where are we now?
If you go back to the lymphocare study that was pre-bendamustine, you had R-CHOP =50%, RCVP =25%, fludarabine-R at around 15% of front line therapy.
I think R-CVP is virtually gone at this point. If BR > R-CHOP > R-CVP in terms of efficacy, I think BR = R-CVP in terms of tolerability. At this point, I do not envision recommending R-CVP except in pretty unusual circumstances. I think BR has also captured a significant portion of the R-CHOP and even the fludarabine based regimens.
I would wager if lymphocare was repeated today it would be BR = 60%, R-CHOP = 30%. You may appropriately read my bias for BR in this post. It is totally fair to ask when would you ever give R-CHOP. I would not disagree strongly with anyone who were to recommend R-CHOP. I still want to see Rummel's paper and I also want to see the initial results of the "Bright" study which was a US version of the same study (ASH 2012?). If that one shows BR to be so much better - then R-CHOP has an even steeper hill to climb.
Occasionally (20-30% at 10 years) follicular lymphoma takes on a bad attitude and "transforms" into Diffuse Large B Cell Lymphoma (DLBCL). Sometimes you have a biopsy in one place, but you suspect based on scans there is a transformation elsewhere. In these cases R-CHOP is completely appropriate as we think the adriamycin is completely appropriate for the transformed disease. Some other researchers are still sticking by R-CHOP since it has such a long track record and perhaps there will be some long term issues with bendamustine that we don't know about yet. This is definitely possible.
I could probably switch gears now and talk about the PRIMA study for rituximab maintenance but I've spent too long on this post already and I would rather get it out there and leave that topic for another day.
If we go back to the lymphocare study as a starting point, back in 2004-2009 timeframe the main rituxan containing chemotherapy regimen choices were R-CHOP > R-CVP > R-Fludarabine.
With the genius of drug regimen naming, I should point out that R-CVP and R-CHOP are essentially the same drugs (vincristine is also called oncovin) with the latter having the addition of hydroxydoxorubicin which we call adriamycin - aargh. R-CVP is mild but the addition of adriamycin is a pretty big deal for most patients. It causes you to lose your hair, it can possibly weaken your heart muscle, and it can increase the likelihood of long term damage to your bone marrow. These regimens were built out of the early chemotherapy days when "combination chemo" was all the rage. The idea was that if one drug was good, four must be better.
To my knowledge there was never a CVP vs CHOP study sans rituxan (though if an reader knows of one, I will be happy to update my post). In the pre-rituximab era CVP became the regimen you used when you had less extensive disease or you were worried about someone being too frail for CHOP. CHOP was the "big guns" and used quite extensively to maximize remission duration.
CVP and CHOP both got a good "facelift" with the addition of rituxan which joined the picture in the very late 1990's. Rituxan was the perfect drug to add to these combinations since it had essentially no toxicity on the bone marrow, was very effective on it's own, and seemed to make chemo work better. There were several studies in which the design was chemo +/- rituxan - I highlight two:
R-CVP versus R-CHOP versus R-FM as first-line therapy for advanced-stage follicular lymphoma: Final results of FOLL05 trial from the Fondazione Italiana Linfomi (FIL).
This study showed that at three years 2/3 of patients who got R-CHOP were still disease free while < 1/2 of patients who got R-CVP were controlled.
The biggest change in treatment compared to the lymphocare study was the introduction of Bendamustine. This drug has a wild history. If you are looking for a cure for insomnia, watch my video on history of cancer treatment which starts with the bombing of Bari Italy and the startling birthplace of chemotherapy. Bendamustine was hanging around East Germany for many years. It wasn't until the fall of the Berlin Wall that folks in the West discovered what the East Germans had been using for decades. I don't think it is an exaggeration to say that this drug has turned FL on it's head.
The change began with impressive results seen in smaller studies of patients with relapsed disease. Even after patients had had multiple prior regimens, bendamustine came in and demonstrated pretty impressive results.
The biggest change though was the head to head matchup of BR vs R-CHOP done by the German group lead by Rummel.
The change began with impressive results seen in smaller studies of patients with relapsed disease. Even after patients had had multiple prior regimens, bendamustine came in and demonstrated pretty impressive results.
The biggest change though was the head to head matchup of BR vs R-CHOP done by the German group lead by Rummel.
Bendamustine plus rituximab (B-R) versus CHOP plus rituximab (CHOP-R) as first-line treatment in patients with indolent and mantle cell lymphomas (MCL): Updated results from the StiL NHL1 study.
This turned a lot of heads when it was initially presented at ASH in 2009 because it showed that BR was at least equivalent to R-CHOP. Since BR is considerably easier on the patient (no hair loss, no cardiac risk, perhaps less heme side effects) - it would be the equivalent of a mid weight boxer taking a heavy weight boxer into the late rounds for a tie match.
Perhaps more shocking however was the updated ASCO presentation this last June when it captured the plenary session for demonstrating a convincingly superior "progression free survival" (time that patient is alive without recurrence of lymphoma). By a wide margin (average 3 years vs 6 years) BR kept patients disease away for longer with less side effects.
I have heard an interesting criticism of the study that the R-CHOP arm performed far worse in this study than virtually all other studies in FL. Perhaps the magnitude of the difference is smaller than what is reported, but in a randomized study, you would expect BR to end up underperforming as well so I am not sure how to interpret that observation.
So where are we now?
If you go back to the lymphocare study that was pre-bendamustine, you had R-CHOP =50%, RCVP =25%, fludarabine-R at around 15% of front line therapy.
I think R-CVP is virtually gone at this point. If BR > R-CHOP > R-CVP in terms of efficacy, I think BR = R-CVP in terms of tolerability. At this point, I do not envision recommending R-CVP except in pretty unusual circumstances. I think BR has also captured a significant portion of the R-CHOP and even the fludarabine based regimens.
I would wager if lymphocare was repeated today it would be BR = 60%, R-CHOP = 30%. You may appropriately read my bias for BR in this post. It is totally fair to ask when would you ever give R-CHOP. I would not disagree strongly with anyone who were to recommend R-CHOP. I still want to see Rummel's paper and I also want to see the initial results of the "Bright" study which was a US version of the same study (ASH 2012?). If that one shows BR to be so much better - then R-CHOP has an even steeper hill to climb.
Occasionally (20-30% at 10 years) follicular lymphoma takes on a bad attitude and "transforms" into Diffuse Large B Cell Lymphoma (DLBCL). Sometimes you have a biopsy in one place, but you suspect based on scans there is a transformation elsewhere. In these cases R-CHOP is completely appropriate as we think the adriamycin is completely appropriate for the transformed disease. Some other researchers are still sticking by R-CHOP since it has such a long track record and perhaps there will be some long term issues with bendamustine that we don't know about yet. This is definitely possible.
I could probably switch gears now and talk about the PRIMA study for rituximab maintenance but I've spent too long on this post already and I would rather get it out there and leave that topic for another day.