Waldenstrom's macroglobulinemia: What the math tells us
Caveats: after writing this, I realized that many parts are wrong. For example, I think resistance to Rituxan is simply due to a poor dosing protocol combined with the protective properties in the bone marrow micro-environment. So "variants" of WM referred to below is likely to be "hard to kill" cells. So read this with a grain of salt. When I have more time (and knowledge), I'll come back and update this.
I thought it would be interesting and useful to see what mathematically modeling WM would tell us about the disease so we could understand what is going on. I don't have sufficient information to do this accurately. So this is just my personal "best guess" as to what is going on.
There is a drug AMDdbaa which isn't the real name of the drug, but a placeholder that is there until it is safe to publish the real name. This has to do with the way medical journals work. It is a real drug. I was just asked by a WM researcher to obscure the name.
I am not a WM expert. I am not a math whiz. What follows are my own personal opinions, theories and "best guesses" as to what might be going on that could explain the data that we observe in this disease. I don't have a medical degree, so I had to guess about a lot of the way medicine works.
Along the way, I had this reviewed by doctors both inside and outside the WM field. I had one of the smartest WM researchers review my 9 conclusions. My personal physician thought it was quite reasonable and a fresh approach and didn't argue with any of the points. One of the top WM researchers who also said that the theories were quite reasonable. Many on the WM list, including many old timers and those who are quite technically sophisticated, thought it was very interesting.
But there are also the detractors. Some people (who aren't doctors) claimed that I can't possibly apply the laws of physics and math to a biological system and get the right answer. One person thought that because I am a new patient that this is just my way of "coping" with the disease. This person didn't point out any flaws in my logic. If you can't attack the merits, I guess you have to resort to personal attacks. One person said that my writings would fail me out of early high school biology and chemistry. So I asked him for the most significant point he disagreed with so I could fix it. Just one point. He refused to name one but just told me I should read more about the disease.
The reason I wrote this is because I was inspired by reading "Patient From Hell" (written by Steve Schneider) which points out that doctors are trained to see things a certain way. The book made the point that an intelligent patient who approaches the same problem from a different vantage point can bring a fresh and helpful understanding of medicine. So Steve Schneider was able to use statistics to challenge his physician and actually change the care that he received. Doctors actually changed protocols based on his logic and he writes about it in his book. Schneider had no medical training.
I am not claiming this is rigorous and fits all the data. It is more intuitive logic of how things work. The conclusions are based on very simple mathematical principles that are generally applicable to physical systems. For example, if a drug is applied at a consistent dose and frequency which causes the rate of cell death to be slightly higher than the chance of division, then the cell population will diminish at a fixed percentage rate over time until it reaches zero. It does not stabilize at a fixed, non-zero equilibrium! Stable non-zero equilibriums happen when the killing mechanism doesn't act directly on the generator, e.g., for a fixed number of WM cells, your IgM level stabilizes when the IgM level reaches a point where the rate of production (proportional to the # of WM cells) is equal to the death rate of the IgM itself. Such stable non-zero equilibriums cannot occur when you are shifting the odds of a WM cell itself because there is only one time constant. The only stable equilibrium when you shift the odds in your favor (even by a tiny bit) is zero. This is critical because it means that as long as you shift the odds so that WM decreases even a little over time, the stable equilibrium point is total annihilation rather than a reduced level of the disease. That means if your drug stops working, it is because you are uncovering a variant that isn't responding to the disease, not because the drug just shifted an equilibrium point downwards.
My theories could be totally wrong. I know a lot of people will disagree. But at least this provides explanations that fit the observations. If there are other explanations that I haven't thought of, please let me know. If you have your own version of this page with different theories that are supported by the math, please let me know. I'd love to compare! But please don't attack this saying I'm wrong unless you can prove I am and can tell me what it should be. If you have a better theory, please share it.
But mainly I hope to stimulate some creative thinking by writing this page. Even if I'm dead wrong about everything, there may be some nuggets here that could inspire someone smarter than I am in a new direction that may lead to a better understanding of this disease.
At least one prominent WM researcher agrees with all of my conclusions. Here they are:
- WM grows because of a delicate probability balance: the odds of cell division are only a tiny bit higher than the odds of cell death without division. The good news is that it doesn't take much of a change in your body chemistry to shift the odds in your favor. The bad news is that it doesn't take much to shift the odds in WM's favor, causing it to grow 5 times faster or more. Unfortunately, we don't know what those changes are that shift the odds.... is it your cholesterol level? something else?
- The risk of creating a drug-resistant WM mutation from treatments is extremely minimal. Even if a mutation of a WM cell happened because of chemo or another treatment, because of WM's slow growth rate (doubles in a year), it would likely take more than 40 years for such a mutation to be clinically significant (reach 1 billion cells). What is more likely to happen is that the chemo could create a very fast growing cancer that could kill you. But such a risk is proportional to the number of chemo treatments, rather than the specific timing of the treatments (see #9). That's because the chemo risk is independent of the disease burden since chemo risks mutating all your cells, not just the WM cells. The bottom line is you are much more likely to create a different kind of cancer from treatment than to "create" a "drug-resistant WM variant" from your WM cells. To minimize such risk, minimize the total amount of chemo agents.
- If your Hct and Hgb look good, but your IgMs are continuing to increase over time, it more likely means that you killed the WM living in your bone marrow but missed a WM variant in some other part of your body (e.g., your lymph nodes). It is less likely that the increased IgM is due to your WM cells exponentially increasing their IgM output over time. WM cells are able to change their rate of production over time (e.g., if BLyS increases per Ansell's work), but it is still constrained over a finite range. Also, it might take 1.5 years after a Rituxan treatment to see the maximum effect in your Hgb. This is not caused by the drug finally "kicking in" after 1.5 years. More likely, the Rituxan killed the WM that was taking up space almost immediately after your treatment, but that it took your good cells 1.3 years to regenerate. So measuring the Hgb is measuring your disease burden very indirectly and the time constant depends on the rate of good cell production and can be very long for that reason.
- Most people have at least two WM variants (variant being defined with respect to drug efficacy; it could be the same genetic WM, just in different places in the body or it could be more than one cell line; Dr. Pliarski found that 15% of patients had more than 1 cell line (each of which preferred a different place in the body) which I learned after reading this and confirmed my hypothesis). Monotherapy treatments almost never produce a complete remission. In other words, with the current crop of drugs, you should expect it will take at least 2 or more drugs to get an improvement of two orders of magnitude or more. Drugs that are proven in cell line testing to be synergistic should (absent proven drug interaction side effects) be used in parallel in order to achieve the maximum impact for a given dose.
- If your drug regimen was working and progressively stops working (i.e., has less and less effect on each treatment), it is not likely that the WM or your body became drug-resistant to that drug. It is much more likely that it is because you've mostly eliminated the variant of WM that was responsive to your treatment and uncovered a variant that was not responsive to your original treatment. Even though your IgMs are not going down as rapidly, you should continue to treat the variant that was responsive by extrapolating the time so that the variant you did attack is relatively small compared with the variant remaining. Then try a different drug. Don't keep trying the same drug and expect a different result. This is true on maintenance therapy as well. If your maintenance "stops working", switch drugs. Conversely, if a drug "failed" before, it may "succeed" later. For example, if you have two variants and variant A is 90% of your disease and variant B is 10% of your total disease, using a drug that attacks the B variant will appear to completely fail since it is masked by the A variant. Once you reduce that A variant by 90%, then that same drug that appeared to fail before, will now appear to succeed. Or suppose the A variant is 70% and your B variant is 30%. Suppose you have a drug that reduces the B variant by 50% on each dose. From your measurement vantage point, you'd see a 15% decrease on the first treatment, a 7% decrease on the second treatment, and the other treatments will have non-measurable impacts (due to the noise). So even though your treatment was highly successful, it will "appear" to your doctor that each time you are treated, it is having less and less of an impact when in reality, it is still killing half the cells on each treatment!
- The reason for not treating WM sooner than later is the fear of irreversible side effects of the treatments, e.g., people have died suddenly from chemo treatments of their WM or have developed other more deadly diseases. If the treatments produced no long term irreversible side effects, there would be no reason to delay treatment. So the reason doctors wait in WM is because the treatments have the potential to be worse than the disease. If you are asymptomatic, they won't treat you. Once you develop symptoms that impact the quality of your life, this balance changes. However, the balance also changes if there are drugs available that can't kill you and produce mostly reversible side effects that are acceptable to you. The bottom line is that the more the disease is affecting you, the wider the range of drugs you should consider. Why would you consider a chemo drug if you haven't tried the safer drugs first? After all, doctors cannot predict which drugs will and will not work with accuracy. If they could do that, then you'd never read about patients who were prescribed drug X and found it had no effect. So it's basically trial and error at this point because our understanding are limited. Try the safest drugs first.
- There currently isn't a good way for scientists to know in advance what treatments will work on you although there are some general guidelines, e.g., Velcade works better on some locations in the body and Rituxan will only work on WM cells expressing CD20. Because there are WM variants, if a particular treatment regimen "stops working", rather than give up, it's important to try a different treatment, e.g., if rituxan no longer works, try velcade, rad001, etc. A drug that stops working, may work later (see #4).
- Try one drug at a time to see if it has an effect. on your IgM or Hgb. That will minimize side effects. There is no point in taking a drug where you can't measure an effect. In general, if a particular drug isn't helping on its own, it's less likely to provide synergy with another drug (exceptions are Perifosine and AMDdbaa which are WM cell mobilizers). Start off with the drugs with the least side effects (e.g., Rituxan, Perifosine, RAD001, Velcade, etc.) for one or two doses to see which have an effect on your WM.
- Treatments in rapid fire at the beginning of treatment are unlikely to be beneficial, i.e., giving 4 cycles of chemo where the cycle time is 3 weeks or less. Fast treatment reduces side effects caused by the presence of WM over time. In general your WM side effects are roughly proportional to your WM burden * time (with some exceptions, e.g., PN which seems to happen with low IgM more than high IgM). However, since WM is slow moving, the reduced WM side effects aren't much of a benefit if the time delays we are talking about are measured in weeks or even a few months. So whether you add an extra week or two between cycles and extend a treatment by a month or two is immaterial from a WM point of view (assuming your symptoms are relatively mild and reversible, e.g., anemia), but it probably helps your good cells to recover between treatments more and thus minimizes side effects of the treatments themselves. From a disease point of view, at the end of the day, all that matters is the number of treatments, and not the specific timing, e.g., unless there is a significant synergistic effect from drug overlap (which can occur for a long time constant drug which is being administered at a frequency that is faster than a few time constants), if you are doing 8 3-week cycles, that that will produce exactly the same net benefit at the end of the cycles as 8 4-week cycles where you are adding an extra week between each cycle. The big benefit here is that by decreasing the dosing frequency, you can avoid (or at least reduce the chance of) serious side effects (such as irreversible PN in the case of Velcade) while having exactly the same total "impact" on your disease. The reason you get a standard dosing schedule is because that is what is "proven to work" in clinical trials so doctors can be sued if they deviate from accepted practice or simply not get paid by the insurance company for doing something non-standard. Or they simply do not want to risk experimenting with an unknown regimen compared to a proven regimen that doesn't work in you. Unfortunately, this is not always in the patient's best interest, e.g., by decreasing the dosing frequency by a factor of 2, a leading WM researcher was able to mostly avoid PN problems caused by Velcade while still providing outstanding patient benefit on a "net benefit per dose" basis. The bottom line is that for WM, in general, adding an extra week or two between cycles is much more likely going to help you (by avoiding side effects of the drug) more than hurt you.
- If your WM stabilizes or goes down, it is much more likely the case that this is due to some changes in your macroenvironment rather than some changes that happened to the WM cells themselves. If your WM is going down year after year without treatment, it is because something in your body shifted to change the delicate odds that allow WM to grow. It is not because your WM cell genetics have been "changed" in some way by the treatment. You have the same WM genetics; it is the environment you changed outside the WM cell that is curing you. It is likely to be a shift in body chemistry or maybe your NK cells are recognizing your WM cells as evil. If we could only see the similarity between patients whose WM is spontaneously going down without treatment, we might have a very effective treatment. For example, maybe people who live in hot climates never get WM. Or maybe people on a certain diet obtain spontaneous remissions. The odds are so close in WM (nearly 50-50) that it doesn't take much to shift them in your favor.
- Triage your options: come up with an order that you will try the therapies that exist. The order should be based on efficacy against where your WM is (for example, I have a lot in my lymph nodes as well as bone marrow), and long-term risk potential (e.g., of developing another more deadly cancer). Keep in mind the following 3 factors: (1) before you evaluate the success or failure of a given treatment, be aware that it may be effective more than a year after you complete all cycles (e.g., some patients keep improving for more than a year after their last Rituxan because their body chemistry has changed), (2) drugs given in isolation may be completely ineffective whereas the same drugs given in combination may be effective, (3) a treatment may semi-permanently alter something in your body's biology or blood chemistry that can cause the WM to grow slower or faster after the treatment is completed. This effect may be more significant than the treatment itself is. For example, with Velcade, the WM typically will grow at either the same rate or a faster rate after treatment. The opposite is true for Rituxan, but not always. In one patient, for example, his WM is now growing five times faster his second round of Rituxan treatment than it was after his first round (he waited 1 year between treatments; each treatment was 1x/week for 4 weeks). So those WM growth-rate related "side effects" should be considered in your treatment decision and not just the side effects that the drug manufacturer tells you about or the immediate efficacy. In fact, it may be in the future, we'll be giving drugs not for their direct impact on the WM cells, but for their impact on your blood chemistry, which will shift the odds in your favor so your WM decreases, i.e., an indirect attack on your cancer.
- It's possible that in the future, treatments may rely more on shifting the odds slightly over time with medicines that are mild and act to shift the odds slightly in our favor and are taken on a continuous basis, rather than shifting the odds a lot over a short dosage time in an attempt to maximize the number killed per dose.
- One person has been on Velcade for 5 months and her IgM dropped from 1100 to 250 and her oncologist says she has a 80% to 90% remission. But look at the math. She probably had about a billion WM cells at 1100. She dropped to 227 million cells after treatment. That's still a lot of cells and if your WM doubles every year, all you've bought yourself is 2 years before your IgM is back where it was before. That's is why complete remissions are only temporary unless the growth rate of WM is reversed post treatment. Each treatment just buys you time. The amount of time you buy is proportional to the effectiveness of the treatment and inversely proportional to the post-treatment growth rate of your WM (which can be faster or slower than what it was before). If the post treatment growth rate is negative, you win.
Here are some things we know:
- For the disease to "show up" in tests, you need a disease burden of about 1 billion WM cells. This is pretty interesting because if there aren't WM stem cells and assuming your WM is growing at 50% per year (like mine is), then it takes 51 years to hit that number. I thought that was pretty interesting because I'm exactly 51 years old. That means that if there isn't a WM stem cell, and WM grows at a roughly constant rate, I was infected with one bad WM cell at a very early age, i.e., nearly the time I was born.
- WM growth varies per individual and can vary in time. Some people can remain stable for 6 months, then change their rate to double every years. A few will double in just 3 months. However, most WM patients will less than double in 1 year.
- Doctors treat based on clinical trial protocols that work. This is for insurance reasons as well as scientific reasons. That's why you get the same dose (per body weight or surface area) as the next person. It's also why the dosing schedule is so fixed. So if you ask to be treated once every 4 weeks instead of once every 3 weeks, your doctor will refuse, even though you can mathematically prove it makes no difference whatsoever in the outcome.
- If there were to be a WM stem cell, it would be reasonable for it to have a 3 day life and produce 10 "bad" WM cells before dying or dividing. The aggregate stem cell growth rate could be very slow or it could double in aggregate in as short as 3 weeks time, i.e., in aggregate, stem cells can be stable or grow very quickly.
- We don't know if there are precursor cells that create more new WM cells or whether WM is solely due to the division of existing WM cells. For sure it is the latter because we can induce WM in mice using just the WM cells.
- Although WM grows slowly, the cells themselves multiply in 2 to 3 days in vitro. Therefore, their chance of dying before dividing is only slightly less than their chance of dividing. That is why WM grows slowly. In other words, a given WM cell might have a 51% chance of dividing and a 49% chance of dying before dividing. Not all cells will have exactly the same odds; this is just an average. So in this example, WM would grow at a rate of 2% per 3 days compounded. So that would be 21% a month. We know WM grows a lot slower than that so the chance of a cell dividing before dying is more than 50% but (for most people) less than 51%. All we have to do is just shift that probability a little and we kill the disease!!!
- Rituxan kills all B-cells which express CD20 equally, but is differential in that it only affects these CD20 B-cells. It has a long half-life (up to 400 hours), but it has been spotted in the body for up to a year or more after treatment.
- Sometimes, people get their best response 6 months after Rituxan treatment, which is beyond 4 half lives of the drug, so that is very odd since the drug is gone by then.
- Several people report that their IgMs have been in steady decline since their last Rituxan treatment 7 years ago.
- The growth rate of the WM cells depends on external factors (inside your body but outside of the WM cell); depending on your body chemistry, your WM will grow at different rates and the rates even within a person can vary over time, sometimes going up and sometimes going down. We know it isn't solely due to your "variant" of WM because when the same cell line is implanted in different mice, it grows at different rates.
- The WM cells that produce the IgM mostly hide in the bone marrow where they are much tougher to kill. There are also a lot in the lymph nodes. Chasing them from the bone marrow into the bloodstream by using stem cell mobilizers helps a lot (synergistic effects have been confirmed).
- The half life of IgM is about 5 days. So if you are using IgM as a proxy for your disease, you need to wait at least 15 days or more before seeing the impact of your treatment.
- Most drugs disappear in <72 hours. Rituxan is a counterexample; it can still be found in your body a year later, but the half life is 16 days so it's virtually all gone in a year.
- Stem cell mobilizers such as AMDdbaa act within 2 hours on the stem cells and the stem cells return home less than 24 hours later. But the WM cells are chased out too! Max mobilization of WM cells out of the marrow occurs after a week of daily AMDdbaa treatments. When the WM cells are in the peripheral blood, they are easier to kill which is why AMDdbaa has a synergistic effect with WM drugs.
- At six months at 1x/week on velcade at 1.6mg, 30% get minor neuropathy that is reversible. Dosing at 1.0 mg/m2 is not enough to have a difference in the cell odds. So changing the dosing frequency is the answer to prevent side effects. It takes about 10 days for normal cells to recover. 1.3mg/m2 is the normal dose. Giving it once every 10 days should (theoretically) minimize any side effects (PN being the most serious).
- Mutations are just random imperfect duplicates. The probability of mutation is proportional to # of cells * time. So the longer you've had WM and the more infected you are, the greater the chance of having mutations. So it is very much smarter to treat WM as soon as possible rather than waiting. Waiting just allows you to create more variants which is going to make it harder to kill. That's why treatments can stabilize at a lower level and not get better. you're killing one variant and the other one is growing. so it's like a parabola. Best chances are to treat early and aggressively to minimize mutations. The downside to treatment: you can die on your first dose of chemo (it does happen). So that's the argument against treating early.
- WM growth is exponential because the rate of growth is proportional to the # of cells unless there is some non-linear limiting factor (e.g., WM only grows in the bone marrow and there is finite space available in the marrow). Exponential growth is a hockey stick upwards. But if you change the odds so you kill more than you make, it isn't a mirror image; the rate decreases quickly and then slows down and asymptotically approaches zero. So a drug that changes the odds in our favor should have a response curve that is steeper earlier on, then gets progressively less. That's exactly what is observed in practice; you quickly reduce and then stabilize at a certain level. So the math explains it. Basically, if you have a fixed drug dosage amount and frequency, your disease decreases at a fixed % rate per unit time. The only reason it will not go to zero is if you are killing just one variant and there is more than one variant.
- It takes about six weeks, for the blood in the body to theoretically completely change. That is, in six weeks the body is capable of making an amount of blood equal to the body's content. This isn't significant at all for the points below, but it does suggest a "recovery" time constant of 6 weeks for the body. So a fast time constant drug applied every 6 weeks should be relatively safe from a long term toxicity effect point of view. However, the normal life of a red blood cell is 120 days (17 weeks). So you only make it at the faster rate if there is a need to.
Based on that knowledge, here are some reasonable theories that explain most of the things we see in WM patients.
Consider yourself lucky: WM is actually relatively easy to cure compared
with other cancers
WM grows slowly, not because it is a lumbering giant that is always moving
forward at a constant rate. It survives only because of a very delicate
balance: the chance of division is only a tiny bit higher than the
rate of cell death; less than 1%! What that means is all we have to do to cure WM is
to SLIGHTLY reduce the odds of a cell dividing or slightly increase the odds of
a cell dying or do both. We only need to shift the odds just a little in our
favor. The more we shift the odds, the faster we can kill all the
WM cells. We do not have to kill WM quickly (although that minimizes chance
of neuopathy and the chance of a drug-resistant mutation). All we have to do is
shift the odds slightly in our favor. That is relatively easy to do
(compared to other diseases)
because the odds are so damn close to even already! So even though WM is
seemingly incurable today, the fact that it is slowly moving due to this
delicate balance makes it one of the easiest to kill. It's still very hard, but
it is far easier than trying to kill a fast moving cancer where you'd have to
shift the odds much more in your favor. A constant, little nudge in the right
direction is all that it should take to change this disease from a slowly
growing disease to one that is slowly shrinking.
WM is incurable because we don't have drugs that are yet proven safe
enough to use it for long enough to kill WM
Suppose you get a complete remission after 1 year. That might result
in a two orders of magnitude drop to1e7 cells instead of 1e9. Then if you continue to treat, you'd need to
treat at the same rate for more than 4 years after you got a complete remission
to kill all the cells. Nobody does this. So since you never kill all the cells,
they come back. It shouldn't be a surprise at all. If you just cut the disease
down by a factor of 100 so it is very hard to measure, stop treatment, and it
doubles every year, it will be back where it was in 7 years. If you can get a
factor of 10 net reduction every year, it would take you 9 years to fully kill
the disease. Each treatment basically buys you a fixed amount of time.
Each treatment buys you time. The frequency of treatment doesn't affect
the outcome. It is just the total number of treatments.
Suppose each treatment reduces your disease burden by 20%. Is it smarter to take
them as close together as you can as early as you can? From a mathematical view,
the answer is no. If you take 4 treatments per year, whether you have all 4
bunched at the beginning of the year or you spread them out over time, or even
have them all at the end of the year, the result is exactly the same. The point
is this: each time you have a treatment, you are simply buying time credits. The
more treatments, the more time you buy. It is the total number of treatments
that matter. So therefore, you should ideally space out the treatments over time
to allow your body to get a chance to recover. The notion that not treating the
disease "hard and fast" at the start isn't supported by the math. You do run the
risk of a higher chance of mutations however. So if you want to minimize
mutations, then you clearly want to front-load the treatments. However, if you
got a mutation, it would probably take at least 50 years or so to be measurable, so this
shouldn't a concern. The bottom line is that treating in rapid fire order at the
very start is mathematically equivalent to spreading things out over time. At
the end of the day, the disease burden is exactly the same.
So the primary advantage to treating rapid fire at the beginning is that you have an opportunity to minimize side effects of the WM such as PN since the side effects are proportional to time and disease burden (this is generally the case, although for PN, it oddly affect people more who have low IgM's than high IgM). So if you want to reduce your chance of getting PN, stroke, etc., then front loading the treatment does that and that's a good enough reason not to significantly delay or stretch out the treatment.
But the math doesn't support the argument that treating aggressively when you do decide to treat will "minimize mutations." That's a valid argument for fast growing cancers, but not slow moving cancers like WM. Even the fastest growing WM variant can grow at 16 times per year so it would take at least 7.5 years for a mutation to appear. But that's the worst case. More likely, for a mutation growing at 2X/yr or less, it would take at least 29.9 years for such a mutation to "show up."
There is not just one variant of WM; we can prove that there are at least
3 different WM variants and probably more. Most people seem to have at least 2
variants.
If there were a single variant of WM, then for any drug that is capable of
causing a decrease in IgM, we should expect WM to continually decrease to zero
(unless the body itself is neutralizing the drug in some way which could happen,
but I think is really rare especially with Rituxan and Velcade). However,
complete remissions are very rare for a single drug treatment. So it's highly
likely that most people have at least 2 primary variants of WM where a given drug
(except for a chemo drug) will only be effective at shifting the odds for one
variant. This explains why dual drug therapies such as CPR and Velcade+Rituxan
are capable of producing complete remissions (CR), while single drug therapies
rarely do. This shouldn't be surprising. There are many strains of TB for example,
and you have to use the right drug for the right strain and a given drug will
not be effective on all strains. So there might be 12 different WM variants. Rituxan may target 8 of them. Velcade might target 6 of them. If you have all
12, you'll never get a CR with just one drug. But there are only two variants, then you are
guaranteed to get a CR if you use both. That doesn't happen in all cases, so
that means that there must be at least 3 WM variants! If you are unlucky, you
will need at least 3 drugs or more to get to a CR (unless you use a chemo drug).
If you are lucky, you'll need only 2. Since CR's are rare for single drug
treatment protocols, it means that most people have developed at least 2
variants.
WM variants can be physical variants, not genetic variants
Don't assume that all WM variants are genetically different. They certainly
probably are for sure. For example, the most common cell defect, the 6q
deletion, is only found in 16% of WM patients (38% in one study at the
University of Toronto).
When cancer cells divide, they tend to divide imperfectly, leading to lots of
genetic variations. Eventually the cancer has countless faces, with a limitless
variety of antigens that need to be targeted by antibodies. That is why, for
example, vaccines
against cancer have been elusive.
So you should feel lucky if one drug killed all your WM.
In addition to genetic variations, there are also physical variants such as the location, e.g., the same genetic WM might be easy for drug A to kill in the peripheral blood, but hard for the exact same drug to reach the same genetic WM that is hiding in the marrow or in a tumor. Or vice versa. So you might have to use drug B, or drug A in combination with drug B, to hit the tumor or bone marrow.
Also, some WM cells may express CD20 stronger than others. Rituxan is more likely to bind to the WM cells with a strong CD20 signal. The other WM cells may be "unreachable" by Rituxan, even though they express CD20.
For mAb antibody treatments like Rituxan, a large number of WM cells can be unreachable for the following reasons (see Monoclonal Antibodies for details):
- they don't express the antigen (e.g., CD20)
- they don't express it in a concentration high enough for the mAb
- the blood flow to the tumor may be insufficient
- high interstitial pressure within the tumor can prevent the passive monoclonal antibodies from binding
A study of a drug that binds to CD20 found that 98% chance of binding in the blood, 96% chance of binding in the marrow, and 60% in the lymph node.
The point is that whenever I refer to a WM variant, please don't assume that it is always just a genetic difference. It might be WM that is in a different place in the body that could be the "variant" or any of the bullet points above.
And don't assume that drugs that work on a WM variant do so at the same rate. For example, a given drug may kill WM in the peripheral blood at a much faster rate than in the bone marrow or in a solid WM tumor. So watch for a double time constant in the data. You might see the numbers drop quickly to Level 1 and then drop slowly thereafter, rather than stabilizing. That would mean you are hitting two different variants and each variant has a different kill rate. And if you use a drug and it goes down and then starts going up again, it's probably because you killed the one variant and now you can see a new variant you never saw before (it was too low to see) that this drug doesn't affect (and not because the drug decided to stop working).
It could be that every WM patient has WM cells that are genetically the same and all variations are due to other factors. I don't know what the data shows on this, but my guess is that the genetic variations are the strongest in terms of sensitivity to drugs, so my guess is that most people have two or more genetic WM variants in their body (since monotherapy seldom produces a CR).
People aren't generally refractory to the drugs in WM even though it may
appear so
If you take Rituxan and it stops working, you probably got rid of all your WM
that had high expression of CD20 that was easy to bind to. The reason your WM is still there could be because you still have a lot
of WM without (or with low) CD20 (or WM hidden in places that Rituxan doesn't
efficiently reach), not that you are immune to Rituxan. Or it could be that Rituxan is great on peripheral blood WM, but acts very slowly on WM in the bone
marrow (i.e., might be two different time constantly of action for the same drug
depending on where the WM is) and that all you are doing is getting rid of the
peripheral blood WM. However, there is lots of data showing Rituxan kills cells
in your marrow (which explains why Hgb improves), so that was just an example. Same
reasoning with Velcade.
A lot of people will just get a 50% reduction on Velcade alone. That just means
that they have another WM variant that isn't sensitive enough to Velcade to make
a statistical difference. If you are lucky (as some people are), Velcade and Rituxan might be
sufficient to cover all your variants.
Increasing the dosing frequency will affect how quickly you stabilize at
your new lower level; it will not affect what that level is as long as you keep
the rate high enough to still decrease the IgMs
It's a mathematical fact that as long as your dose + frequency is still
sufficient to cause your disease to go down, that all you are doing by
increasing the frequency of dosing is accelerating the rate at which you get to
equilibrium. It will not change the final value (which is zero in all cases
except if the drug is being neutralized before it gets to the cell).
Therefore, changing the
dosing frequency just affects how quickly you get there. So you can reduce or
eliminate side effects just by decreasing the dosing frequency. This will just
lengthen your treatment time; it will not affect the final outcome. The final
outcome is that all WM cells sensitive to your drug are reduced to a very small
value. The reason your IgM may still be high at that point is because you only
killed one of the variants. So your stable point will be some new IgM value; it
will be 0 if there is only one variant that you killed. It will be some reduced
value if you have >1 variant. However, you cannot lower the dosing frequency to
be arbitrarily low. It must still be sufficient to tip the odds in your favor
and cause the IgMs to go down. If you decrease the frequency too much, you won't
have a sufficient shift in the odds and your WM will grow.
There are probably two time constants for measuring the impact of a drug
WM cells are in the blood as well as the marrow. A given drug is typically
going to be able to kill WM cells in the blood faster than in the bone marrow.
So there are two different kill rates. However, in practice, most of the IgM is
produced by the WM in the marrow so that is the only time constant that matters.
But in general, if we could measure it, we'd probably see both time constants
but this would be quite hard to measure as the kill rate in the blood is
probably pretty quick and there probably isn't much of it.
Each drug that isn't able to reduce your IgM by 20% should be tested after
hitting equilibrium with AMDdbaa to see if a lower IgM value can be obtained.
This isn't patient specific but it should be tried in people for each WM drug to
see whether the combination is sufficiently synergistic to justify the time and
expoense.
Each drug used for WM is going to have a different kill effect for peripheral
blood WM vs. WM hiding in the bone marrow. So suppose Velcade only shifts the
odds for WM cells in the periphery. We'd get to a lower IgM value and things
would just sit there. The key test to make is then give AMDdbaa for 7 days, and
then give three velcade doses on day 3, 6, 9 and stop. See if this lowers the
IgM at all. I suspect it won't because people have gotten 50% IgM reductions
with Velcade alone suggesting that it is indeed reaching and killing WM cells in
the marrow (otherwise, we wouldn't have gotten more than a 10% reduction). There
is a much faster kill rate with AMDdbaa. But in most cases, it should not change
the final outcome; just how fast you get there. However, for a very targeted drug that that
didn't work before, AMDdbaa might make a huge difference because the reason for
the low effect might be inability to impact the cells in the marrow. So it can
open up more treatment options for us, .e.g., enable a drug for the variant of
WM that you have that was previously untreatable.
If you get a bad side effect to a drug, you may not have to stop using it;
you should first try to take it less often
Try changing the dosing frequency. For example, at 2X/week dosing frequency, 50%
of Velcade patients got PN very quickly. But if you reduce the dosing to
1X/week, only a small % got PN and it happened after 6 months. Also, new
generations of proteosome inhibitors from the Velcade family show no neuropathy
in early trials.
As long as the dosing frequency is sufficient to cause your WM to decrease, changing the dosing frequency shouldn't change your outcome...it just delays it. So unless you are in a big hurray, reducing the frequency can allow you to keep using a drug that works for you while minimizing side effects.
The same is true for single infusions as well. People who cannot tolerate Rituxan infused at the normal rate (e.g., blood pressure drops dangerously low), do just fine when infused over a 24 hour period.
Treating aggressively and sooner is best
I know this goes against what they tell you, but think the question is all in
the choice of drugs. If you only can choose between drugs with a significant
chance of bad permanent side effects, then waiting (for a better drug to become
available) might be better. But if you can see what happens if you treat with
drugs which are minimally likely to have permanent side effects, then I think
that changes the advice. If you pick from safe
drugs that can't kill you and have minimal reversible side effects, customize the
treatment to the patient to only use drugs that work (testing one at a time),
and minimize side effects as outlined on this page,
then I don't think there is any value whatsoever in waiting. All you are doing by waiting is
increasing your chances of getting PN and increasing your chances of additional
WM drug-resistant mutations (or WM growing in places that are hard to reach).
That is a bad thing. So I think the newer drugs that we have now in clinical
trials (or soon will be in clinical trials) changes things. That's my opinion.
For example, many people on the talk list complain about peripheral neuropathy
or lowered IgG or IgA. So you are risking permanent by not getting
treated. And by the time your PN symptoms appear, the irreversible damage has
been done. You can have problems for the rest of your life. Why risk it?
The other thing is that if your disease is stable, that is the easiest time to kill it! If your disease is stable, it means that the birth rate=death rate. All you have to do is shift the odds by a tiny tiny bit and you're on your way to health. And there is very little risk factor because all the drugs either don't work or improve your numbers. They don't make things worse for your disease (but they can create permanent side effects). However, they could change your macroenvironment for the worse and shift the odds in favor of WM and cause a WM that didn't grow to be one that grows slowly. So if you don't have a drug that can kill the smoldering WM you have, then what's the point? But if you do have an effective drug with few side effects, I'd argue that the smoldering stage is going to be the easiest time to kill it and if there are reasonable drugs available, you should try them.
To keep your remission, keep treating it after it appears to be gone for
at least as long as it took you to get there; then switch to a lower dosing
frequency
I think a lot of people make the mistake of stopping treatment when they get
to a complete remission. WM is good at hiding out. Unless you are looking at a
LOT of cells, you might still have 1 cell in 10,000 that are bad that can regrow
again. So the trick is to keep treating the WM even after it appears gone. This
is like why you keep giving antibiotics even after all your symptoms are gone.
Same deal. So suppose you have 10M WM cells in your bone marrow and suppose your
WM doubles every year. I'm just making this up. You then give a magic drug that
causes your WM to decrease by a factor of 10 every 6 months. Well, in just 1
year, your WM is at 1% of where it was. And in two years, it is 1/10,000 of
where it was. So your IgM's are back to normal. But there were 10M cells. Now
there are just 1,000 cells. Too small for you to find. It's like trying to find
a needle in a haystack. If your WM doubles every year, in 10 years you are at 1M
cells. So your WM has "returned." What you really want to do is beat back your
WM to a low level. So if you killed it in X months, you want to continue dosing
at the same rate for at least another X months to kill the stuff you cannot see
and knock down the counts. Keeping the total cell count low reduces mutation
chances. Then you just continue on using a maintenance dosage that will keep the
disease in check. This is going to be less than the normal dosage rate. You're
trying to find the dosage frequency at which the disease growth is exactly
neutralized, i.e., where the drug provides just enough of a benefit to
counteract the growth of the disease so as to keep it in check. That minimizes
toxicity, reduces cost, and reduces the amount of time you have to spend at the
hospital. So you should measure your rate of disease progression rate before
taking any drugs, and then measure the initial rate of IgM decline for each drug
at a given dosing frequency. So for example, suppose your WM is growing at 30%
per year. You do 3 cycles of Velcade (3 doses per 4 week cycle), and your IgM
drops 30%. That will neutralize the growth rate (approximately). Since you
probably have more than 1 WM variant, that means you've actually killed more
than 30% of your variant. Let's be conservative and call it 30%. So it means
your "maintenance dosage" is 3 cycles of Velcade spread out over the year, i.e.,
9 standard doses. So you'd then give the standard Velcade dose 9 times, evenly
spread out through the year. This minimizes any chance of toxicity or side
effects, while still keeping the disease in check. You'd need to the maintenance
with each drug that was required to kill your disease. This is why if you can
kill your disease without the chemo drugs, that is highly desirable. So a
Rituxan-Velcade option is highly preferable to a CPR treatment because the
former is a much more pleasant and safer maintenance experience. Another way to
calculate your maintenance dose is using the method in the next item.
If you measure your response rate to each drug with 3 data points, you can
tell what % of your WM cells respond to the drug
You can project how well each drug is going to work for you with just 3 data
points: the starting IgM point and two others. You can then fit an exponential
decay to these points and extrapolate out your end point. A non-zero IgM
endpoint means that you have more than one WM variant and this drug isn't
attacking the other variant (or is attacking it with a much slower time
constant). Generally, only one time constant is observed for a given drug (i.e.,
people tend to exponentially approach a stable point rather than having two
different slopes) suggesting that drugs tend to work only on one WM variant. If
a drug worked differentially on two WM variants, we'd observe a quick
exponential decline, and then a slower exponential decline from that point
forward.
Methods to calculate your maintenance dose
By characterizing the IgM decrease curve for two different dosing
rates, you can then extrapolate what your maintenance dosing frequency should
be. For example, suppose your IgM went down 20% per month at 2 doses a month and
went down by 5% per month at 1 dose a month. Solving the equations, we find that
the WM growth rate is 10%. So the drug if used alone at 2x/mo was a 30% decrease
if there was no WM growth. At 1 dose/mo, the drug gives a 15% decrease, which
becomes 5% due to the growth. Therefore, if 1 dose/mo gives a 15% decrease, a
2/3 dose per month will give a 10% decrease which offsets the WM growth rate.
The point is, from two different dosing rates (or from the progression rate
without treatment and with treatment) we can calculate the maintenance
dosing frequency. Another crude way to do it is look at the minimum time to
progression in clinical studies and then apply that dose at that rate. Even
better is to measure your own time to progression and dose at that rate. So if 5
treatments gave a 1 year time to progression, you'd give 5 doses a year
(generally spread out evenly over the year).
There is a good reason why your IgM can suddenly go from completely
"normal" to "doubling every year"
Say your WM disease is now doubling every year and it wasn't before. You may
wonder...wow....I never had it before and now it is "suddenly" doubling every
year?? It was stable for every and now it takes off??!?! Well, it was probably
always doubling for years and years but you never knew it because the numbers
were so low compared to normal IgM number. Once the numbers became high enough
that the monoclonal IgM was being produced at a rate that shows up on your
tests, then you started to notice it. So it looks like you had nothing and then
your disease just "decided" to take off even though it was taking off from years
ago. It's like viewing a chart of world population. It looks like a hockey
stick. Your WM resulted from one bad cell many many years ago. The number of
those cells has grown exponentially over time, so that you now take notice.
There might also be some synergistic, positive feedback effects that are going on that once WM starts growing, it makes it grow faster...sort of like global warming and positive feedback as described in Inconvenient Truth. But I have no evidence that that is true. It's at least growing exponentially and that's plenty bad.
If you want to avoid a maintenance dosing schedule, then you should keep
taking the same medication until you can reduce the disease by 100X or more
Let's say you have a pretty significant disease that is growing 2x in a
year. You knock it down via treatments over a 1 year time period and now it is
unmeasurable. Can you stop completely with the drugs? That's a bad idea. Your
body has a gazillion cells. They are going to take a drop of blood to analyze.
Suppose they could look at every single cell in that drop of blood and see if
there is a WM cell there. I pretty seriously doubt that they can do that, but
say they can look at every single cell in a drop of blood and find nothing.
Great. That just mean that the average number of WM cells per drop of blood is
less than 1. How many drops of blood do you think you have in your body? The
point is that your body could still have a huge number of WM cells still in the
blood even though your sample is completely clean. And there can be a huge
number of WM cells hiding out in a tumor or a lymph node or in your tissues that
won't show up in a BMB or blood test. Or there can be WM stem cells that we
don't know about and have such a subtle genetic difference that no test is going
to pick it up.
So if you truly want to avoid the disease "returning", you have to kill it all. That's nearly impossible to do because you might have a variant of WM hiding out somewhere that doesn't show up on the tests that only drug A will cure. WM hiding in a tumor may not show up at all in a BMB, but it will in an IgM test, but only if the level of production is high enough to see. You won't know to cure it until it shows up on your blood tests. So there isn't much you can do about the WM you don't know about.
For the WM you do know about, if you want to avoid a maintenance dosing schedule, then you need to knock the disease down by a factor of 100 or more. A factor of 100 of course doesn't eliminate the disease; it just makes it 100 times smaller so if your WM is slowly growing, a factor of 100 will buy you a lot of time. The faster your disease is growing and the more time you want your remission, the greater you have to knock it down. So suppose it took you a year to knock it down by a factor of 100 so that it went from measurable, to virtually unmeasurable (assuming you can still detect a 100X decrease in concentration). You'd still need to continue treatment for an additional period to get the reduction you need. The amount of time depends on how fast your WM is growing compared to how fast you are killing it.
Your actual numbers will differ, but the point here is to make the calculation in terms of how long it takes for a given treatment to reduce the disease burden by a factor of 2. You look at how many years you have before you die (which is probably a reasonable remission time!), and you look at how fast your WM is growing. Armed with those numbers, you can calculate how much longer you have to keep treating at the current treatment rate until you have the remission time that you are looking for. That calculation is only for WM that is reachable by the drug you are using. You have to do it for each drug separately.
The other options is simply switch to a maintenance dose when you hit a plateau again, using the calculation of drug kill rate compared to disease growth rate.
Be careful though. There might be more than 1 time constant. Each WM variant might have a different growth rate, e.g., tumors may grow at one rate, bone marrow WM might grow at a different rate. And they could be genetically different too. So look at the WM your drug is targeting. That may not be so easy...you have to do it while your WM is still easily measurable so you can do a kill with the drug and then see how fast it comes back. However, other variants will be growing at the same time, so picking the growth rate of the WM you are affecting may be very hard to see distinctly.
How a drug can still working after 7 years
People who have Rituxan report that their IgM is still dropping 7 years
after they were last treated. The Rituxan is long gone after about a year. So
how can that happen? The only explanation is that your body chemistry has
permanently shifted by the treatments and that has shifted the very delicate
odds that WM requires to grow just slightly in the negative direction. Now if we
could just find out what is "different" about a patient after Rituxan vs.
before, we'd have a very potent weapon.
Some people think that the immune system kicks in and that explains the reduction. Personally, I find that way too hard to believe. WM is a slow moving cancer and the immune system is a quick kill system that normally knocks out an invader in a matter of days. So the time constants just don't make sense; if the immune system "kicked in", you wouldn't see a steady decline over 7 years; you'd see an instant decline in days. Secondly, I find it also hard to believe that the immune system would recognize a fellow B-cell as an antigen to attack. That is like attacking yourself. I may be wrong, but I don't think this ever happens. In AIDS, it is a foreign virus that attacks the immune system. And in auto-immune diseases, it is the immune system that attacks other parts of the body, not the immune system itself. There is no case I'm aware of where the immune system attacks and kills B-cells.
Why the max response for Rituxan occurs around 2 months
The half life of Rituxan is less than 400 hours (16 days). The kill rate should decline at an exponential rate and become nearly zero after
about 4 time constants, i.e., 64 days. So I'd expect that the IgMs would hit
their nadir at 2 months after the last infusion, and then start going up from
there. And that's what we observe.
Why Rituxan can last for months or years after infusion
The Rituxan kills the B cells and lymphoplasmacytic cells and not the plasma
cells that produce the IgM. Therefore, you're killing the "young" malignant
cells, and not the "adult" cells. So the adults (which produce the IgM) die off
of old age and they aren't replaced with new adults since you killed all the
young. Therefore, the population of adults decreases which is why your IgM
decreases.
Here's the more "technical" explanation from Treon's paper, Extended rituximab therapy in Waldenström's macroglobulinemia which says the same thing (but is just harder to understand):
An intriguing finding in this study was the late response activity, as evidenced by the median time to best response of 17 (range 3–32) months. Continued declines in serum IgM beyond 1 year were also noted among some patients in the study by Dimopoulos et al. [13
]. One possibility for this finding is that rituximab may differentially target members of the malignant clone in WM. It is well recognized that in WM, the malignant clone encompasses mature B cells, lymphoplasmacytic cells and plasma cells [2
Why there is IgM flare caused by Rituxan?
This one continues to baffle me (and everyone else). Mathematically, the only way you get IgM flare is if you
increase the # of WM cells (i.e., by shifting the odds), you increase the death
time constant of IgM (so the cells live longer), or you increase the rate that the WM cells are making IgM
or do some combination. I'd guess that the Rituxan is temporarily altering the
death time constant of the IgM, but that is just a guess. It's probably not
changing the # of WM cells since you eventually return back to your normal rate
and there would be no way for the body to remember what cell count that was so I
think that is pretty unlikely.
Another possibility is some sort of temporary positive feedback loop, e.g., the
rapid release of IgM makes it more probable that more IgM gets release (although
it seems unlikely that that would be the mechanism). Another
explanation is when the rituxan kills the WM cells, they release IgM out right
before dying. I haven't spent enough time thinking about this though.
One really important point is that the flare is the same monoclonal. So the flare is not your body making antibodies against the mouse antibody of the Rituxan.
Are the Rituxan infusion related reactions (chills, hives) caused by an
allergic reaction to the mouse antibodies in the Rituxan?
Guy ask me:
"But, are the hives as a result of the murine component of the rituximab, or are they as a result of rapid cell kill of the WM cells in the circulation (and bone marrow to a lesser extent) and release of chemical mediators (cytokines) in the bloodstream ? - which is thought to maybe cause the chills and shakes to some extent. All hypothetical and not thoroughly proven I don't think."
No, I think it isn't due to the murine component. You get the same reaction in Humax so it's release of cytokines which is good!
There is some evidence that HuMax infusion reactions could be more severe
than with Rituxan according to this excellent article:
http://www.clltopics.org/ClinTrials/SmarterMonoclonal.htm
this is why they keep the dosing rate REALLY slow for the first 300mg.
Not clear about the delay onset side effects of humax v. rituxan. We'll see.
Seems to be no delayed onset neutropenia. See:
http://www.clltopics.org/ClinTrials/HarveyIsBack.htm
See also:
http://www.clltopics.org/Rituxan/RituxanRoadBlock.htm for side effects
not generally mentioned for Rituxan, e.g., delayed onset neutropenia, hepatitis,
and viral reactivation.
How to get rid of IgM flare on Rituxan
Since we don't know what causes the flare, it's hard to know how to get rid
of it.
Could be that HuMax may not cause any flare at all. This would be VERY interesting to try.
When Rituxan eliminates B-cells, BLyS goes up which is perhaps the body's means of compensating for the loss. BLyS has been proven to increase the rate of IgM production and reduce WM cell death. There are probably other factors operating as well. So while the Rituxan is trying to kill the WM cells, your body is manufacturing proteins to help them live.
Perhaps if the Rituxan is administered at low frequency in the beginning, we can eliminate Rituxan flare. The standard dosing schedule that everyone uses is 1x/wk for 4 weeks. Then you wait for a while and do it again.
After all...what's the rush to treatment? Before you got treated, they were giving you nothing. Now, all of a sudden, they turn on a firehose. What's the hurry? Unlike a drug like Velcade, mAb therapy is completely independent of concentration. The only thing increasing the dosing schedule or the amount in a single dose is doing for you is increasing the rate at which we improve; it cannot not change the final endpoint at all. So if you aren't in a hurry, giving Rituxan once a month may be a super safe thing to do. Rituxan maintenance has been done at 1 dose every 3 months or 1 4 dose cycle every 6 months. So why not start with a maintenance cycle and see if you can eliminate the flare?
Suppose you instead do 1x/mo for the first 2 months and then gradually increase the rate? What would happen? Does anyone know? Perhaps not since doctors seldom deviate from the protocol tried in the clinical trials.
When the dosing rate should be increased
Normally, just changing the dosing rate just impacts the time to a given
endpoint. But in cases where the new drugs act synergistically with a previously
administered long time constant drug, the dosing rate does matter for maximum
synergistic effect. For example, if you are giving CPR treatment, the cytoxan
and prednisone are both short lived, but the Rituxan is not and it overlaps the
doses. By giving doses every 3 weeks, you allow time for the body to recover,
but maximize the amount of Rituxan that can combine with the cytoxan and
predisone in order to maximize the kill rate. So by giving doses before the
Rituxan dissipates, there is less total drugs input into your body required for
a given kill amount. In that case, more frequent dosing to take advantage of
known synergies are reasonable.
Why IgM counts go back to normal within 30 days of a plasma pheresis
There is a lot of IgM in the tissues. So if you remove the IgM from the
blood, you are going to get a quick diffusion from the tissues into the blood to
equalize the IgM concentration. Then there is a longer time constant effect
caused by the the half life of the IgM itself (5 days). One should expect that
the IgM rises to a constant level at the time constant determined by the death
rate of the IgM itself. Therefore, one should expect 63% of full value in 5
days, 86% of full value in 10 days and 98% of full value in 20 days. That's why
everyone gets to full level after PP in 30 days (to account for varying time
constants). The reason it stabilizes is because the rate WM cells are emitting
(which is proportional to the number of WM cells) becomes equal to the death
rate of the cells (which is proportional to the # of IgM cells). So there is
always a stable equilibrium for everyone based on the # of WM cells they have,
and you'll always get there in less than 30 days, no matter how good your PP is.
Why IgM is stable
IgM is relatively stable because the number of WM cells is relatively stable
(WM grows slowly)
and so they produce IgM at an extremely constant rate. The natural lifetime of
IgM then sets the stable equilibrium point. Since the lifetime of IgM doesn't
change, measuring the IgM level is a great proxy to measuring the full impact of
WM.
How we know that the IgM production rate is stable and is proportional to
the number of WM cells
I've said that IgM level is a great proxy for how much disease you have.
Some people think that because their Hct and Hgb are fine even though their IgM is climbing to new highs every year that they are fine. They think that their IgM is just increasing because maybe it is not being scavenged as efficiently or because their WM cells are just producing more IgM. Both of those theories are just incredibly statistically unlikely. Let's examine each one.
Say their IgM is increasing because it isn't being scavenged as effectively. Actually, the IgMs just die on their own. But let's say there is something that shifted the IgM death constant so that the IgMs live longer than they should. All this does is change the stable equilibrium point. It cannot cause a consistent super linear increase over time in the number of IgM (which is what is observed). It would just be a one time event.
The other hypothesis is that the existing WM cells are simply making more IgM. That might be a believable hypothesis if there is a step function in the number of IgM that happens after treatment because the microenvironment has changed (e.g., IgM production increases in the presence of stromal cells) since that would be the only way to affect the huge number of cells that exist. They certainly aren't going to independently decide on their own to do this at the same time! And any such increase would be relatively small... and probably not measurable since it would swamped by the noise. But that is not what is observed. What is observed is an superlinear increase in the number of IgM over time even though their blood counts are improving. So you'd have to believe that your WM cells are decreasing (if you believe Hgb is a good proxy for the the number of WM cells which it isn't) and my IgM is increasing at a superlinear rate because my fewer WM cells are overproducing IgM like crazy? Not only that, but the rate of IgM production from each cell is increasing at an exponential rate!!! How is that possible? It isn't. Cell chemistry makes it impossible. A single cell cannot exponentially increase its output over time and then have it's offspring copies be able to make more and more IgM per unit time at an exponentially increasing rate, i.e., each offspring makes exponentially (over time) more IgM than their parents did. That's biologically impossible from a single cell. It just ain't going to happen.
So we are left with only one alternative: the rate of IgM production is constant over time for a given WM variant and the half life is as well. Therefore, the superlinear increase in IgM is due to a superlinear increase in the number of WM cells. Which means measuring IgM is a great way to see how bad your disease burden is.
Maintenance is always a good idea, even if you have a complete remission
Because we don't know whether there is a WM stem cell, even if you have a
complete remission, you should periodically check to make sure the disease truly
is gone. But if it is very low it is hard to measure. A low dosage rate
maintenance therapy, designed specifically for you, as outlined above is the
best protection against a WM stem cell causing your disease to return.
WM cells may be most vulnerable to being killed when they are dividing
I speculate that the WM cell is probably most vulnerable to being killed is
when it is dividing. There is probably not a lot you can do to take advantage of
it. And I'm just guessing.
You want to test a new drug at the highest safest dose and dosing rate to
see if it has any effect
If you have too low a dose or too low a dosing rate, you can have no impact with
a drug that would otherwise be effective. So to test a drug on yourself to see
if it changes your IgM, try the standard recommended dose, at an aggressive
frequency. If there is an effect, you can then lower the dosing frequency to
minimize side effects so long as you are seeing IgM's still decline.
If your IgMs are going down on their own, you may not need maintenance
Your IgM is already known to be decreasing (either because you've shifted
your body chemistry or because you still have the drug in your system), then you
don't need maintenance. The whole point of maintenance is to neutralize the
untreated growth rate of your WM. See above on how to calculate the correct
maintenance dosage for you.
For treatment, you'd want to use the most aggressive dose you can that
doesn't cause side effects since this eliminates the disease the fastest which
reduces time for mutation.
W&W seems like a bad idea from an eventual treatment point of view. You
just give the disease lots of time to develop drug resistant strains. The rate
of mutation shouldn't be any higher during treatment vs. non-treatment. It is just proportional to the number of cells and time.
So the sooner you treat, the better. However, if your treatments are "dangerous"
(e.g., chemo drugs) this doesn't apply.
If there are WM stem cells, they probably don't play a primary role
Nobody knows if there are WM stem cells. As long as the stem cells are small
in number, we may never know the answer to that question Unless there is a magic
bullet that can reach and kill all WM cells, then finding a WM stem cells may be
nearly impossible.
There was nothing in the math that can prove or disprove the existence of a WM stem cell. However, if you wipe out a WM variant and it grows back at a linear rate at first, instead of an exponential rate, then that would be hard to explain without a WM stem cell.
Suppose that WM's growth is primarily due to stem cells. Then we'd have a situation like the IgM - plasma pheresis situation, i.e., if you increase the kill rate of the offspring, you'll reach a new lower equilibrium point. So that "could" be what is happening with WM, e.g., when you treat with Velcade and get a 50% remission, it "could" be because the kill rate is now increased so there is a lower equilibrium IgM point. But if that were true, then changing the dose frequency of the Velcade should change your equilibrium point. It doesn't. Changing the dose frequency just affect how fast you get to your new level, not what the new level is. We know that because when people switch to Velcade maintenance (a lower frequency of dosing), the IgM levels don't quickly stabilize at a new higher value.
Another key point is that if you removed the Velcade, if WM was primarily generated by stem cells, then you'd see an exponential equilibration back to the level before you applied the Velcade and it would equilibrate very quickly, e.g., in way less than 30 days (which is well beyond the WM cell life which would determine the time constant of recovery). But we don't see such a rapid recovery and the response curve is exponentially increasing (typically at the same rate as the disease normally progresses).
So therefore, this disease is primarily determined by the WM cells themselves, although the stem cells could play a role. That re-affirms my hypothesis that there are several variants of WM and that if you just get a 50% IgM reduction, it has to be because you've eliminated one variant; without a stem cell involvement, the only stable equilibrium is zero (by contrast, the IgM case has a non-zero equilibrium point because the death mechanism is proportional to the offspring and not the rate of production).
You could argue that the reason that the IgM doesn't rapidly rise is because the stem cells are only growing the WM population at a very slow rate which would explain the slow recovery time after treatment. But if that were true, then then the kill rate would be virtually 100%. Velcade can produce 50% reductions in a month, but then it stabilizes. If the stem cells are growing IgM at a slow rate, that isn't sufficient to "stop" the Velcade momentum. The only way you'd equilibrate at 50% is if you had a pretty serious push back. And then when you removed the Velcade, you should see a very rapid rise back to where you were. Since you see none of this, the role of stem cells is relatively minor, other than as a source of possible disease which makes maintenance critical.
The other way to deduce that this is probably not stem cell related is to observe the growth rate. If it is exponential, the disease is self-replicating. However, I admit that the stem cells could also be growing exponentially as well as the B-cells. We just don't know and it would be hard to find out.
If we could engineer a mAb that only attacked the WM cell, and we gave it for long enough so we can't detect any WM cells, then we'd know. We'd kill almost everything and then remove the drug. If the disease returned, we'd surmise there are a lot of WM stem cells since that is the only way the disease could come back so quickly.
Another way to do it is count the number of WM cells in a BMB. Then take 2 more samples a month apart and see whether it is growing linearly or exponentially. It's virtually guaranteed that the number of WM cells you have in your sample is different than the number of WM stem cells. So if your sample is in fact growing at an exponential rate, then that's the primary source. The reason for killing all the WM cells is that there might be a small number of stem cells and the only way to see if that is true is first to eliminate the WM cells so their numbers don't swamp out the stem cell numbers.
Due to the multiple variants of WM, it is very difficult to detect synergy
between drugs if you test in humans (which is why they test for synergy using
cell lines)
If you get a 30% reduction in IgM with Drug A alone, and a 30% reduction with
drug B alone, and then you combine them and get a 70% reduction, then there is
some sort of synergistic effect going on. Similarly, if the rate of improvement
is more than that predicted by the combination of the drugs, that also
demonstrates synergy. But there still could be synergy going on, even if you
don't have a combined improvement. For example, suppose you use both drugs and
you still only get a 30% reduction. The drugs could both be killing the same
variant of WM that is 30% of your WM total. So if the kill rate is faster than
predicted by the combination, then you have synergy, even though the outcome is
exactly the same final level of reduction.
Let's say drugs A and B each target and kill a different WM variant and you have 2 variants. So you apply A and get a 50% reduction, then apply drug B and get a 100% reduction. There is no synergy at all here that can be proven (unless you measure the rate and the rate with both is faster than you'd predict from each drug combined assuming no synergy). So it's not like you killed "more" of the same type of WM with the 2 drugs. It's much more likely that you killed two different variants of the disease.
The place you might get clear synergy is with a WM cell mobilizer. That should enhance the effectiveness of a given dosage frequency, allowing you to progress faster (or use a lower dosing frequency).
Synergy between drugs may allow you also to reach new variants. If there are 3 WM variants and drug A only affects WM#1 and drug B affects only WM#2, then it's possible that with both in your system, you could affect WM#3. But that remains to be shown and the experiment to prove this is tricky to set up so it proves the point.
The key thing here is designing an experiment that proves there is synergy. I think the only way to do that is within the same patient, you time shift the dosing to be in parallel vs. series and then see if you can see a differential in the rate of decrease as well as the final endpoint achieved when the dosing is changed. The final endpoint is probably the most interesting and that is easy to test. Apply the drugs so that there is no time overlap until you get stability. Then time shift your dosing schedule so the drugs are given together. Does it further reduce the IgM level? If there is a synergistic kill such the drugs combined can kill a cell, while either drug alone is insufficient, then you'll see an additional IgM drop when the drugs are administered together.
The way synergy is typically done today is with cell lines. So you can see what happens with drug A alone, drug B alone, and drug A+B combination. You are looking typically at the depth of the response. if A+B response is > max(A, B), it means there is synergy because since it is a single cell line, it eliminates the "variants" argument. So this is a pretty good indicator but it would be nice to do the differential study in people to confirm it works the same way in humans.
My personal preference as far as using multiple drugs is to start with a single drug, wait for stability, then add the next drug to see if there is additional benefit (a lower stability point).
Hitting the WM cell more often probably doesn't improve the kill % of each
dose
I think there may be a belief among some people that if you hit WM more
frequently, before the cells have had a chance to recover from your first "wave"
of attack, that you improve your kill rate. Therefore, for a given number of
doses, if that theory were true, the final IgM would be lower than if the doses
were spread out further in time. That might in fact be true, but I think the
effect would be pretty minimal. The dosing rate of Velcade in Treon's trial was
2X per week. In Ghobrials, it was half the frequency. But I'm pretty sure that
the results were the same. Therefore, all the faster dose rate does is reduce
the time required to hit a stable point, it doesn't lower that point. This is
advantageous as the quicker you get rid of WM, the better. But the flip side is
2X/week caused PN in 50% of the patients. So the benefit of the quicker rate is
outweighed by the sometimes non-reversible PN. Bottom line wisdom for me is that
there is no good reason to be in such a big hurry to kill this slow moving
cancer. After all, the agreed protocol is generally watch and wait. So why then
suddenly go from an attack rate of zero for years and then suddenly start off
with guns blazing? A rate of 1X/week can get to a stable point pretty quickly
(very quickly relative to normal disease progression time) and it eliminates the
side effects for most people.
Hitting a disease "hard" (e.g., high dosage frequency) is more
likely to cause
a drug resistant mutation rather than prevent one, but the reduction in total
cell numbers creates a net risk reduction so hitting it hard is the right thing
to do in order to reduce mutation probabilities
One thing you want to avoid is for the treatment itself to increase the
probability of a mutation that might be drug resistant. Doctors like to treat
aggressively at first to reduce the WM count since mutation likelihood is
proportional to the number of cells times time. So that makes sense, I agree. But suppose we treat with Velcade at the maximum dosing frequency. So we might actually be hitting cells
that have not yet fully recovered from previous dose. You'd expect such cells to
be a real mess beforehand, and in even worse shape afterwards. Therefore, if
they do divide, they probably are more likely to make more "mistakes" than a
relatively healthy cell. That's because DNA replication is amazingly accurate in
cells. Disturbing the cell is never going to make the copying process MORE
accurate. That would be so unlikely because cell duplication is nearly flawless
already (DNA replication is actually more flawless than the bit error rate on your typical disk drive). So
disturbing that mechanism is much more likely to result in errors (mutations).
Therefore, hitting the WM cells while they are "down" from the first dose is
going to increase the chance of creating a disease resistant mutation. However,
you have to weigh the additional risk of a successful mutation against the time
and number of cells. Invariably, time and number of cells wins because this is
typically an order of magnitude reduction in probability (e.g., if you get a 90%
reduction). That is why doctors are correct in hitting the disease hard and fast
since that minimizes the number of successful mutations that can occur.
It is extremely unlikely that the treatment itself is causing a
drug-resistant strain of WM to be created
Some people believe that when a treatment has stopped working on a patient,
it is because the disease has "mutated" to become resistant to the drug. I don't
think so. These treatments work so fast (in some cases, a month), that there is
very little time available for a new mutation that just happens to be drug resistant to the
exact drug
you are using to be created and then grow to a size that is measurable. It just
isn't going to happen. Even if you were super unlucky and you did create a
mutation, it would take years to grow to a measurable size. Much more likely is
that in the 8 years you had MGUS, all the mutations happened and grew to a
significant size.
So if you are getting a partial response, it is because you are only hitting one of the mutations that existed before you started (or because the drug has totally stopped working on you because some other part of the body is neutralizing it), not because the disease adapted and became resistant to your treatment. I cannot come up with a mathematical model to explain such a hypothesis. Think it through...you start with 1M WM cells you treat it and you just get a 20% reduction. 80% of the 1M cells didn't change their DNA overnight. If you created a drug resistant mutation, you'd start with a couple of cells that are resistant. Those cells would take years to grow to a size you would be able to measure. And in the mean time, the WM you were trying to kill would be all gone. So you'd never see a profile of just an 20% or 50% etc. reduction. So your WM is NOT morphing on you during treatment.
So this notion of WM "becoming" refractory to your treatment is pure hogwash. Your drug seems to stop working because you killed the WM variant that the drug could shift the odds on. One more supporting piece of evidence that I'm right is that nobody get a CR from Velcade alone. So if the disease becomes "drug resistant" then it means that in 100% of the patients, that a magic drug resistant variant was created in just a few months. The odds of that happening are zero. It would be as likely of having 10 people flip a coin and them all getting heads. But the most powerful argument against "your WM became refractory" is, as stated in the previous paragraph, that there is simply no time. WM grows slowly. A mutation that is drug resistant, would take years to grow. So the only possible explanation is that either (1) you are just hitting the WM cells that respond to the drug you are using and that is why your drug seems to "stop" working or (2) something is completely neutralizing your drug before it gets to the WM cell. Option (1) is far more likely as explained in the next point.
It is also really unlikely that your body has developing an immunity to the drug
as the explanation as to why a given drug appears to "stop working"
Drugs like Velcade have been shown to keep doing their thing over and over
in patient after patient. The macroenvironment (your blood chemistry) changes.
There is no resistance to that change. So if there is a resistance to Velcade, it would
have to be in
the target WM cell itself that it has bypassed the mechanism that Velcade
disrupts. But we know that that's impossible from the previous point. So the
only possibility left is the "we just hit that variant that was sensitive to
Velcade." It cannot be that the drug was "neutralized" by the body before it hit
the WM cell because the body doesn't do that as pointed out in the first
sentence. But even if we didn't know that fact, we could still deduce it from
the physics. When Velcade is injected, suppose we had a mythical large army of
super Velcade eaters like PacMan that would gobble up all the Velcade as soon as
it was injected. Well, cells don't gobble that fast, a good part of the Velcade
is going to hit its target because we've injected a huge amount all at once. So
that means that the drug would still be effective, just to a lesser degree. So the observations (plateau) don't fit.
The same argument can be made for Rituxan. I don't think there is any data showing that the body develops antibodies to the Rituxan antibody. I don't think the body would ever do that since it would cause havoc in the immune system if antibodies were cannibalistic. So the Rituxan that is injected are full functioning antibodies when the hit the WM cell. So if the drug isn't working, it's because the Rituxan cannot reach your WM, or your WM expresses CD20 too lightly or not at all.
Bottom line: if you aren't responding with an IgM drop to zeo, you have WM somewhere in your body that simply never was responsive to the drug (or responsive very slowly). I don't think any other explanation fits.
There are drug synergies: cytoxan and Rituxan, AMDdbaa and Velcade, etc.
There are drug combinations that are clearly synergic, where the drugs, when
used together, provide an increased kill rate. So for example, cytoxan and
Rituxan would be synergistic. They both weaken the same cells since cytoxan
isn't cell specific. Such drugs should be used most effectively at the same
time, rather than alone. For a given dose, you maximize the impact. But
combining drugs can sometimes hurt you (drug interaction problems where the drug
overlap happens to destroy really good cells) and sometimes can help you.
Drugs can be combined in serial or parallel
The best way is to test combinations is seeing if the combination resulted in a
deeper kill rate when the drugs were used simultaneously rather than
sequentially.
So to prove synergy, you'd start with the drugs in series and wait till the IgM stabilizes. Then you shift the dosing to use both drugs at the same time (you might have to give one drug earlier if the time to maximum effect is different for the two drugs). Then you see if the IgM goes down further after the change to using the drugs in parallel.
In series is far safer. So unless you you can show an a deeper reduction in disease in parallel, why take the risk? If the combination is known to be safe and synergistic, then that's a great reason to use the drugs in parallel. So look to see if a clinical trial can measure a difference in the same patient when drugs are given in series vs. parallel as in using the test above. If there is no significant faster impact by using the drugs in parallel, then it is silly to do that because you risk unknown interaction problems for no benefit.
If the parallel synergy effect is minimal, using the drugs in an alternating series will cure you just as fast as taking them in parallel, but will minimize the interaction risk. For example, suppose you have drug A that can be given every 7 days, and drug B that can be given every 7 days as well. If you gave them in parallel, you'd simply give both drugs at the same time. If you gave them in series, you'd simply alternate between the drugs every 3.5 days. So the rate is the same, but we've minimized the drug overlap time which should minimize the risk of drug interactions (at the expense of minimizing synergy).
Why your WM starts increasing again after treatment is stopped
There are two reasons for this: 1) you are seeing the growth of the variant
that is left after you killed the one that your drug worked for and 2) unless
you continued using your drug well after it seemed to have stopped working, then
you didn't kill it all and it is just coming back. In general, it's a good idea
that if it took X days to reach stability, then continue to treat it at the same
rate for another X days. But it depends on what level of improvement you got
too. If you only got a slight improvement, you actually have to treat for a lot
longer to really ensure you are reducing the disease. This is because of
"masking effects." For example, suppose WM variant #1 contributes 10 IgM and
variant #2 contributes 100 IgM. If you reduce variant #1 by a factor of 2, it
becomes lost in the noise of IgM #1 and you become "stable." But you really want
to reduce it by a factor of 10 or more. So you need to keep dosing.
Why after a BMT, your WM can return "with a vengance"
If you don't kill all the WM cells in doing a BMT, the WM will return. And
in the process of doing all the drugs you need for a BMT, you have disturbed the
macroenvironment. Sometimes that works in your favor. Other times it doesn't and
it tips the odds of growth just a teeny little bit in favor of the WM. Just a
tiny change in the odds can make WM look like it came back with a vengance. This
is because the odds are so close to 50-50. Change them by just 1% and it can
create an aggressively growing WM. A 51%-49% is a 2% net growth per division
(which happens every 3 days). So that's a growth rate of 22% per month (your WM
will increase by a factor of 10 per year)!!!
Why your Hgb can be going up and responding to treatment at the same time
your IgM is steadily increasing
Michael L's IgM level have been going up steadily since diagnosis, but
his Hgb responds to Rituxan quite well (his hgb goes up after treatment and then down
like an upside down parabola whereas his IgM stabilizes during rituxan and then
continues to climb after treatment. So what's going on here? How can we explain
it?
My guess is that the Rituxan must be clearing out the WM from his bone marrow so it has room to make Hgb. But he might have another variant of WM that seems to have more of an affinity for hiding out in his lymph nodes or wherever (or that has no specific hiding place) that are not being affected (as much) by the Rituxan and are just merrily increasing at an exponential rate. So just because your hgb is looking good, you can't just say you are cured. If your monoclonal IgMs are increasing, it means your WM is increasing.
So he sees his Hgb improve because he clears out the WM from the bone marrow. And his total IgM goes down because we've eliminated the WM from the marrow. So over time, the IgM in the marrow comes back and he retreats to restore his Hgb. But in the meantime, the WM in other parts of his body have been steadily advancing. So this explains why the marrow can look stable over time (Hbg constantly getting reset to normal), while the IgMs still increase. And his IgM variation decreases on each Rituxan treatment cycle because the total WM in his marrow is probably reduced by a lot so when we kill it we don't see much of a decrease.
One possibility it is that his "untreatable" WM is in a clump in lymph nodes or in a tumor in his body somewhere. Such a clump of cells, let's assume they are in a spherical shape, could emit IgM proportional to the volume of the sphere, but there is only the surface area of the sphere to attack from a treatment perspective. This makes such clumps very hard to treat. In addition, there is probably a very tiny positive pressure outwards from the center of the clump with all the IgM's going out from the sphere. So a Rituxan mAb who gets close to the sphere is going to be "pushed" away from the pressure and can't get close enough to stick (sort of like my trying to approach someone holding a firehose; i'm not going to get close). I was really guessing at this and then found out I was right (see Immunotherapy for Cancer - Monoclonal Antibodies where it talks about "high interstitial pressure within a tumor" can prevent a mAb from binding).
So I'd guess that mAb therapy, due to the way it works, would be quite a bit less effective against a WM tumor or WM growing in the lymph nodes due to these two effects (surface area and outward pressure). That would explain why the bone marrow WM decreases, while his IgMs are still going up.
How a drug can start working on WM when you tried it before and it didn't
work at all (and also why drugs that worked "before" can seem to "stop" working)
Just because a drug appears to have no effect when you first try it, it might
have an effect later! How can that be? Simple. Suppose you have 3 variants of
WM. Suppose the first two variants are now huge and the third variant is still
really tiny. You try drug #3 and nothing happens. That's because variant #3 is
tiny and so you can't see the impact. A year later you've killed variant #1 and
variant #2, using drug #1 and drug #2 but the disease "returns" and the drugs
that worked before to kill them before don't work now!! Well, that's because
you're trying drug #1 and drug#2 on wm#3. You just assumed you are getting the
same variant that you killed before back and it has come back in a drug
resistant format. But the reality is you killed #1 and #2, and now #3 is large
enough to see. So that drug you thought was useless before is now your main
weapon... you just couldn't see the effect until variants #1 and #2 were reduced
enough, and variant #3 grew enough.
Using Rituxan is actually extremely targeted therapy compared to other WM
treatments
The number of normal B-cells in circulation and in the bone marrow which
express CD20 is actually very small compared to the number of WM cells. That
means that the chance a given Rituxan mAb hits the right target is around 90%. A
given mAb can kill it's prey, and then release and kill another B-cell. So it's
pretty efficient. The half life of Rituxan is around 21 days but longer if it
re-circulates like this.
The best way to treat WM
Ideally, what we want is a monoclonal antibody that just targets the WM cell
without touching anything else. That's
not possible to make the normal way, because animals don't know how to make an
antibody to the WM cell itself.
So we'd have to synthesize it and I don't think we are smart enough to know how
to do this. But if we could, that would be great. Today, we can make "designer
genes" (Codon Devices, British Bio-technology, GeneArt, Blue Heron, DNA2.0). See
How do you like your genes?
Or a WM vaccine would be great too.
But the problem with either approach is the genetic variations usually found in cancer. You'd have to target them all to cure it.
This web page on Helping the Immune System Fight Cancer discusses the problems with mAbs and vaccines for cancer. We aren't there yet.
Someday, radio waves may cure cancer.
That is why existing approach focus on some fundamental pathways that WM relies upon, no matter what the variant. By closing down 2 or 3 pathways, we can probably eliminate this disease in most people.
Treatment summary
Here's a summary:
- measure how fast your wm is growing
- start with one drug. measure the treatment reduction rate.
- If no reduction in disease at max dose and max dose rate, try another drug.
- when it plateaus, keep treating until you've reduced it to about 100X and switch to a maintenance dose (or keep treating at the same rate to achieve a 10+ year remission calculated from the rate of disease progression)
- monitor at least IgM and Hgb since they could be moving in opposite directions
- if IgM is still not normal apply the next drug and repeat. Try to avoid overlapping drugs unless clinical trial show a synergy
- as new, lower levels are achieved, remember to re-try drugs that didn't see to work before since you may not have been able to "see" the currently dominant variant before because it was masked by other variants
Based on all of this, here's what I am thinking of doing and why
I am 10% involved right now. My conclusions are to treat early and aggressively, but back off the dosing frequency to just enough to eliminate side effects so long as the frequency is still sufficient to cause the disease to diminish. I want to characterize each drug in terms of the effect on my IgM levels to see which ones work against my WM and are complementary, rather than duplicative. I should also pick drugs to use that are safe to apply in a maintenance setting.
In my opinion, the only "safe to use for maintenance drugs" that are FDA approved and have been tested in WM are Velcade and Rituxan. Velcade and Rituxan have already proven (in combination) to provide CR's in WM patients with relapsed disease. Velcade as a mono therapy has a 90% response rate. There is now a clinical trial open for newly diagnosed patients. Both Velcade and Rituxan have been used in patients for more than 2 years without problems (I know of one person who has been using Velcade for 4 years). But the seemingly irreversible PN from Velcade is a big problem.
There are also Campath (80% response), Viagra, Gleevec, and statins which can be tried.
Velcade has a very short response time (1 month to see significant drops) where as Rituxan takes longer (2 months or more till maximum response). I would expect that Velcade can reach places (like deep inside tumors) that Rituxan cannot (but I think that this is not true for some reason). And I know people with a 90% IgM decrease (6000 to 600) on monotherapy Velcade after 72 doses (18 21-day cycles)
So first I want to characterize my overall WM growth rate without drugs. That will allow me to calculate my maintenance dose just from observing the kill rate of a given fixed dosing frequency of an effective drug.
Since Velcade acts faster, it makes sense for me to characterize my response to Velcade first (e.g., 1.3mg/m2 dose on day 1 and day 4 followed by a 10 day rest should be much safer than the standard protocol, yet pretty effective; dosing once every 10 days would be even safer) and see what level I can go down to before my IgM stabilizes. This isn't much lower than that causing very fast remissions for patients so I suspect this is sufficient to result in reductions. If it isn't going to 1.6mg/m2 1x/week would be as aggressive as I'd want to try. If there is no response to that, then I give up on the drug.
If there is a response to Velcade, then I'd want to continue to use it at the same rate for at least the same amount of time that it took me to get to stabilization. That should reduce it to about 1/10,000 of where it was before (because when you hit plateau, you've probably reduced the strain you are hitting to 1/100 of what it once was). Then I can switch to a maintenance schedule for that drug as described above since there is no need to continue at a killing dose any longer (although if I want to get off the drugs entirely, I'd keep going at the initial dosing rate for another 2 time constants).
In parallel with that, after the IgM has stabilized from the Velcade (and while I am still on the same dosing frequency as I started), I'd add in Rituxan to see if I can lower the IgM value to zero. This will tell me whether the Rituxan is actually complementary to the Velcade for me since it might have no effect at all (which is possible since the Velcade may be getting the "easy to kill" strain of WM). If it works, then I'll look at the time to CR and continue the Rituxan for at least the same amount of time before switching to maintenance dosage. Then if the two together are not sufficient to reduce my monoclonal IgM to being not measurable, then I will likely skip over:
- Gleevec (imatinib)
- Statins
- Campath (alemtuzumab) which has some really bad hematological side effects
- and viagra
and instead enroll in a trial for RAD001 (Everolimus), Perifosine, SGN70, hsp90, or Enzastaurin and add that to the mix or use them if they are FDA approved in time.
In general, clinical trials are needed to advance what we know about medicine. I am wary of those that put you into a "one size fits all" regimen because they usually specify a fixed "treatment protocol" that is, in many cases, independent of the patient's reactions (however, many will adjust things on a per patient basis, e.g., "continue with this dose until the disease reaches a plateau," but in general, everyone is treated the same, e.g., give 12 cycles of drug A then stop). One size fits all is bad from a patient point of view, but I can see why the investigator likes it. The same dosing frequency isn't appropriate for all patients and the same drugs work differently in different patients. If it takes me 3 months for my IgM to plateau from Velcade, then that is the point I'd want to add Rituxan (because then I can clearly see if Rituxan is additive) and not on some arbitrarily set fixed day after the start of the trial. If everyone reacted the same, it's appropriate to treat people the same. I guess the regimen makes it easier to evaluate the results since everyone had the same treatment. I just don't think it is in the best interests of the patient if the protocol doesn't allow for patient variability in response times or whether they respond at all, e.g., a trial which starts with both drugs simultaneously is missing a key fact as to whether that is due to both drugs or whether the patient was just responsive to one of the drugs.
My Stanford doc doesn't like starting with Velcade due to the incidence of PN and that's a good point. I told her that if you reduce the dosage frequency, the chance of PN is minimized, but she said she hadn't seen the published data on that so is reluctant to agree. In looking at the website, it says only 51% of patients improved if they got neuropathy. So the concern is very valid since in the chance that I get it, I may be stuck with it. However, the statistics were at a high dose frequency of Velcade and I'm in no big hurry. And at the first sign of PN, we can back off the frequency.
A friend of Jeff Atlin has been using Velcade for 4 years on maintenance dosing frequency, so that is conflicting information. Jeff himself got a 90% reduction in IgM on Velcade monotherapy with a minor PN flare and then it went back to normal. Furthermore, if Velcade ate away your nerves every time it was used, I doubt it would get FDA approval or that physicians would ever prescribe it on a maintenance basis. Still the fact that only 51% got improvement of their PN after dose adjustments is concerning. In addition, cases of severe sensory and motor peripheral neuropathy have been reported. However, it could be if you dose at a very low frequency, you never get PN. Would be nice to test this out.
So she thinks I should start with monotherapy Rituxan because that is really safe. She's right about the Velcade. The fact that PN happens at all means that good cells don't recover quickly (and sometimes never). In Ghobrial's study, when people got PN, it seemed like it happened when the cumulative amount of Velcade reached a certain level regardless of the frequency (but you could go a lot longer than just the cumulative amount would suggest). So that suggests that each dose of Velcade might be doing some small amount of permanent damage. So that makes it a problem for a maintenance therapy. However, the newer proteosome inhibitors show no PN in early trials so that is good news.
Treon doesn't like monotherapy Rituxan at all because it never gets to a CR and secondarily because there is risk of IgM flare. So he likes CPR (cytoxan, prednisone, and rituxan) which has no IgM flare and the 3 drugs work synergistically to kill WM. The dose would be 1 cycle per 3 weeks where a cycle is a single dose of Rituxan and Cytoxan and 5 days of Prednisone for up to 8 cycles. However, the cytoxan can have you lose your hair so who knows what other bad things will happen to good cells. What is troubling is that when your hair grows back, it may grow back differently. One has to wonder what other bodily systems are permanently degraded by cytoxan. Furthermore, the conventional wisdom says avoid alkylator agents if you are <65 to reduce the potential of stem cell damage and of DNA damage leading to a new cancer.
Here's Treon's reasoning for CPR for me, with Drs. Ghobrial and Advani in agreement:
The response rates and even attainment of CR is much better with CPR than we ever saw with rituximab monotherapy, and that's without even factoring in the extensive adenopathy that you possess. Velcade is a disappointing drug as monotherapy for extramedullary disease and responses are short lasting. Hence my preference would be CPR followed by maintenance R.
November 17, 2007 addendum
I assumed all WM cells are alike. That's wrong. There are 3 stages to a WM cell and that's important since each divide and die at a different rate, they are present in different proportion in your body, and the treatments can affect only cells in a certain stage of development. You can think of the stages as the cell as is ages, from child, to teenager, to adult.
- B lymphocytes (express CD20)
- plasmacytoid lymphocytes aka lymphoplasmacytoid cells (express CD20)
- plasma cells that produce monoclonal IgM but do not express CD20
The term "lymphoplasmacytic cells" (LPC) refers to all of them. "Plasmacytic" means "having to do with plasma cells, i.e., they all can produce IgM.
We know their properties (life times, division rates, etc.) in vitro; but not
in vivo, e.g., the cell lines can divide in 3 days in vitro.
However, it seems pretty likely to me that WM plasma cells don't divide much, if
at all (since if they did, then Rituxan wouldn't be able to cause IgM to drop
monotonically for years after treatment).
Why Rituxan causes IgM to drop for five years or more after your last
treatment
When you take Rituxan, you are ONLY killing the kids and the teenagers. You
aren't affecting the adults (the plasma cells) that are producing most of the
IgM at all because the plasma cells don't express CD20 anymore.
The adults eventually die off from old age. It's not known how long WM plasma
cells live. Some can live as long as 20 years, but most don't. Most probably
live for 5 to 8 years based on what we are seeing for IgM die off after a
successful Rituxan treatment. So if your treatment got all your WM B-cells,
you'll see your IgMs gradually go down every month for several years after your
last treatment. What you are seeing is the plasma cell population die off from
natural lifetime causes. The adult population dies off of old age because the
Rituxan killed all (or most of) the kids so there are no (or few) new adults
being produced. So you just have a death rate which should be exponentially
increasing because the young were produced at a faster rate because the disease
grows over time. But for most people, the death rate will look very linear due
to the slow growth rates of WM.
Unlike plasma cells which have a very long life, the IgM has a half life of
around 5 days which is why your IgM is stable for a given number of plasma cells
rather than growing without bound; a linear IgM production rate and an
exponential rate of IgM death gives a stable equilibrium for IgM. So IgM is is
fairly fast responding indicator of how well you are doing with respect to those
WM cells that produce IgM.
But the IgM measurement is a "time delayed" measurement of the effectiveness of
the Rituxan therapy. If your Rituxan works perfectly and removes every CD20 WM
cell, only the adult plasma cells are left and they'll die out over time. So
after a successful treatment with Rituxan only, your IgM will not drop
instantly; instead it will nearly linearly decline over time (it will actually
go down slightly faster over time) and go to zero as you approach the "average
lifetime" of your WM plasma cells (which can take years).
So that's why you see what appears to be a "spontaneous remission" within
monotonically decreasing IgM for many years after completing your last Rituxan
treatment.
Rituxan resistance
If you are seeing Rituxan "stop working" it could be because you've picked off
all the easy targets. WM cells, like other diseases, express CD20 at various
intensity levels within the same person. Rituxan isn't great at binding to any
cell with CD20; the stronger the CD20 is expressed, the better. So if you see it
start to become less effective with each treatment, it's likely because you got
the strongly expressed CD20 WM cells, and are not getting the cells that express
less CD20. However, by increasing the concentration of Rituxan to well above the
standard dose (which you can do by taking the standard dose, but just increasing
the frequency of administration), you may be able reach cells that weren't
reachable before. See
Rituximab Dose-Escalation Trial in Chronic Lymphocytic Leukemia -- O’Brien et
al. 19 (8) 2165 -- Journal of Clinical Oncology. You can also try HuMax or
GA101 (see
GSK reveals details of $2bn antibody) which are supposed to bind more
strongly to CD20.
Rituxan dosing
If you are given monotherapy Rituxan, they'll typically give it to you once a
week for 4 weeks. But if you are given Rituxan as part of chemo, e.g., rCHOP or
CPR, then you'll get it once every 3 to 4 weeks. So which is more effective? 4
doses spaced a week apart? or 4 doses spaced a month apart? All things being
equal, assuming your Rituxan dose doesn't saturate all the CD20 sites, it
shouldn't matter. But it could and surprisingly, nobody knows the answer to this
in vivo.
Velcade successes may be short lived
People have reported achieving remarkable reductions in IgM using Velcade.
However, if the drug is stopped, many people find that their IgM levels stop
declining within a couple of months and then quickly return to normal. This
suggests that the Velcade is interfering with IgM production while it is active
rather than causing a large amount of permanent cell death
Related links
Waldenstrom's macroglobulinemia: A new approach
Steve Kirsch home page: my personal home page