Treatment Decisions

The VH gene story emerged late in 1999 together with CD38, FISH a year later and ZAP-70 in 2003. They have been the subject of the Education sessions at ASH and ASCO. There is no excuse for anybody treating CLL not knowing about them. BUT, no clinical trials have yet reported making use of these prognostic markers, though several are being conducted.

We are agreed that there is only one disease, but equally agreed that there is a spectrum of severity from fatal within a year to never causing a day's problem in 35 years. However, it is not a straight gradation, the disease is clustered around two poles. One is always troublesome, requires treatment and shortens lifespan, the other is seldom troublesome more often than not does not require treatment and in most cases does not shorten lifespan. The most secure way of separating these two types is by mutational analysis of the VH genes: those with mutations have an average survival of 25 years, those without mutations have an average survival of 8 years. 17p deletions by FISH virtually only occur in the unmutated subset, and make a bad situation worse, with an average survival of about 2-3 years. 11q deletions by FISH mostly occur in the unmutated subset. Their prognostic impact is not so clear-cut, but in about half the cases they also make a bad situation worse. The other prognostic markers, CD38 and ZAP-70 were introduced in the hope that they would be a surrogate for VH gene mutations which are difficult and expensive to do. Unfortunately, there are two many exceptions for them to be used in this way. In about 30% of cases CD38 and VH genes give different answers. Those that have disparate results for these tests have an average survival of 15 years. ZAP-70 gave the same answer as VH genes in over 90% of cases in the first two papers, but the latest study shows 25% disparity. Again, having both positive is worse than having only one positive.

So, how should we use these tests to design treatment aims.

At present people not in trials are either following the standard NCI criteria, which do not take account of the new markers, or operating on guesswork.

The questions that remain to be answered are these:

Should patients with bad prognostic markers be treated by watch and wait? Nobody knows, but we should do trials aimed at comparing w&w with attempts to cure the disease while it is still of low volume.

When patients need treatment for clinical; reasons, should patients with good prognostic markers be treated in the same way as those with poor prognostic markers. Again we don't know, but we should compare simple and safe treatments like chlorambucil with more effective, but more hazardous treatments like FCR in patients with good prognostic markers.

When patients have 17p deletions should we go through the less toxic treatments like F or FC or FCR first before we go to the more effective but more hazardous treatments like CAMPATH? Again we don't know, but this question should also be addressed by a clinical trial.

Why 6 rounds of treatment?

The idea is to get the maximum response. Even after lymphocytes have
disappeared from the blood there are still at least 1000 million left in the  body.
When they can't be detected in the marrow by looking down the  microscope there
are 100 million left. Even when they cannot be detected by  even the most
sensitive test there are somewhere between 10 and 100 thousand  left.

The fact that 17p or p53 deletions might be resistant to fludarabine is fairly recent information. Some patients with p53 deletions, however, get CRs on FCR. Early relapse is common though. Trials on how to treat p53 deficient patients are just starting. There are a number of studies just beginning. For example there is a joint Dutch/Scandinavian/Polish trial of FC versus FC+Campath about to start in patients who need treatment and have either Unmutated VH or p53 deletions or 11q23 deletions or trisomy 12. Similar trials are under way.

Because these trials are not complete we really do not know the best advice for patients who are p53 deficient. Conventional wisdom says that they will need Campath, or high dose steroids or an allograft, but this is all based on very preliminary data.

When to start treatment? And with what?
 
I have long been an advocate of not starting treatment before it is really  
necessary. But when it is necessary then the sort of treatment you should have  
should be determined by clinical trials. 
 
What are the indications for treatment? The National Cancer Institute has  
issued guidelines on this.
 
One of  the items on the following list must be present: 
1]         Weight loss of more than 10% of body weight in the previous  6 
months;  
2]         Extreme fatigue so that the patient cannot work of perform  usual 
activities;   
3]         Fevers of greater than 100.5°F for at least 2 weeks without 
evidence  of infection;  
4]         Night sweats without evidence of infection. They have to be  
severe. 
5]         Evidence of bone marrow  failure shown by the development of, or 
worsening  of,        anemia or  thrombocytopenia. 
6]         Autoimmune hemolytic  anemia and/or thrombocytopenia poorly 
responsive to  corticosteroids. 
7]         Massive splenomegaly (>6cm  below the costal margin). 
8]         Massive lymph nodes or clusters of nodes (>10cm in  longest 
diameter). 
9]         Increase in lymphocyte count by >50% over two months or 
anticipated  doubling time of <6 months. 
10]       A high lymphocyte count is not in itself an indication for  
treatment. 
If treatment is necessary the bottom line to look at is overall survival.  
Trials are often reported in terms of response rates or progression free  
survivals. Now it is nice to go into remission, and nice to stay in remission a  
long time, but if that means that you don't live as long, it is a bit of a  
pyrrhic victory. 
So what do we know about overall survival. Despite thousands of people  being 
studied there is no evidence that fludarabine as first line treatment  gives 
a better overall survival than chlorambucil. This is despite fludarabine  
having a better response rate and a longer treatment free survival, because  
patients seem to do as just well when they receive fludarabine as second line  
treatment. 
Fludarabine and cyclophosphamide perform better than fludarabine alone in  
phase II trials, but when they go head to head in a Phase III trial, we only  
have the German trial to look at. The interim results of this show an overall  
response rate of 95% for F+C and 84% for F alone. The complete response rate is 
 20.3% compared to 8.6%, the progression free survival is 46 months compared 
to  21 months; but so far there is no difference in the overall  survival. 
So far we have no head to head comparison between FC and FCR, though the  
Germans also have an ongoing trial of this. Recently published in the J Clin  
Oncol are two papers from the MD Anderson on their results with FCR. They report  
an overall response rate of 95% for FCR and a complete response of 70%. These 
 are impressive results, but we are unsure of the mix of patients who were  
treated. This is where a full breakdown of prognostic factors is needed. The  
most important distinguishing factor is the VH gene mutational status. Mutated  
patients seldom die of CLL: unmutated patients usually do. Treatment should 
be  different for the two. If a study contains a majority of mutated patients 
the  results will be a lot better than if it contains a majority of unmutated  
patients. 
Few patients know their mutational status, but ZAP-70 has been suggested  as 
an acceptable surrogate. It is in 77%. Suggestions are made that it a better  
distinguisher than VH genes, but this has not been confirmed. 
FISH tells you about p53 status. People who lack p53 or have mutant p53  need 
to be considered as a separate group. 
In my ideal world, where these tests were available for everybody, I  would 
have everybody in a clinical trial because there are still many things we  
don't know. But it ain't going to happen, so in the real world, if you  definitely 
need treatment, what should you do? 
First, if you are in the p53 group, FCR is not the answer. The drugs  
available are Campath (which is no good for lymph node based disease) and high  dose 
steroids. I recommend that patients get both together, and that  data are 
collected. In the UK we are running a phase II trial looking at  this combination. 
You might also consider having a mini-allograft if you get a  good response 
to these drugs. 
Second if you have unmutated disease, you have a choice. Most people are  
choosing FCR in these circumstances. Despite the high response rate, there are  
no data to suggest that you do better in the long run if you start with FCR 
than  if you build up to it gradually, using it only when you have relapsed after 
 using lesser treatments first.  It has been said that in Oncology practice  
you should use your most effective treatment first, but while that may be true 
 for AML, it is not true for say, prostate cancer, where radical  surgery on 
everybody would lead to significant overtreatment of many  patients with 
unnecessary side effects. So, there is no right or wrong  answer. 
Third, if you have the mutated sort and need treatment, then there is  much 
to be said for not going for FCR even though it is likely to produce a long  
remission. Because the disease has a much longer natural history, there is a  
greater risk of living long enough to get late side effects of infection,  
Richter's syndrome and MDS. Again there is no right or wrong answer. You might  
consider Rituxan plus G-CSF or GM-CSF as first line, or chlorambucil or  
chlorambucil plus Rituxan. 
I'm sorry if this is a bit rambling, it is an attempt to answer several  
questions that have appeared on the list. As always when the right randomized  
clinical trials have not yet been done, there are more questions than  answers. 

Why do different countries have different standards for first line  therapy?
 
It is a question of different philosophies for health care. The cost of  
health care has been increasing at such a rate that governments have had to  limit 
health care spending. If trends continue in the USA the cost of health  care 
will approach 20% of gross domestic product. In Germany and France it  is 
about 12% and the UK 7%. In New Zealand, last time I looked it was about  3.4%.  
I'm not sure of the figure for Canada. So called socialised medicine  costs 
less, but does it deliver? One of the strategies that governments have  adopted 
to slow down health care costs is the introduction of Evidence Based  Medicine. 
The idea is that for a treatment to be funded it has to have been  shown to 
be more effective in a randomized controlled trial (RCT) than the  treatment it 
is replacing. When this policy was introduced it was embarrassing  to find 
that many well established treatments in medicine had never been tested  in an 
RCT. However, it proved impractical to deny them all. You have to start  from 
where you are. A refinement of this policy was to add in health economics.  Not 
only was a new treatment to be more effective, but also cost effective. This  
was measured in QALYs - it stands for quality adjusted life years. 
(see _http://www.evidence-based-medicine.co.uk/ebmfiles/WhatisaQALY.pdf_ 
(http://www.evidence-based-medicine.co.uk/ebmfiles/WhatisaQALY.pdf) . 
 
In other words how much does it cost to buy an extra year of good quality  
life. To the patient any cost would be justified, but those who have to pay for  
it - the tax payer in the UK, the insurer (and hence businesses) in the US - 
set  a price beyond which they think it too expensive. In the UK an acceptable 
cost  of a QALY is about £30,000. (However, the cost of a life saved by 
putting in  crash barriers between carriageways on motorways is £1 million and the 
cost of a  life saved by putting in advanced passenger warning systems on the 
railways that  automatically brakes the train if it passes a red signal light 
is £3  million.)
 
Let's not be naive; whatever people say about equality, some lives are  
regarded as more valuable than others, and there are ways of getting round the  
system. I can't imagine a government minister being denied a particular  
treatment because the cost of a QALY was too high.
 
On the particular question of FCR versus chlorambucil and prednisolone  there 
is a particular problem because of a lack of RCTs. There is absolutely no  
doubt that FCR will produce longer remissions and better remissions than  
chlorambucil and prednisolone, but there are no studies that show that people  live 
longer if their first treatment is FCR rather than chlorambucil. (In fact  
even by combining all the trials that have compared fludarabine first with  
chlorambucil first there is no difference in overall survival.)
 
Because patients with CLL may live for 20 years or more, to do a trial like  
this would take a very long time. By the time it was completed the trial drugs 
 would be out of patent and new agents would probably have become available.  
Because of this physician/scientists have adopted surrogate endpoints. Not 
"How  long will a patient live?", but, "How long will a remission last?" or "How 
many  will go into complete remission?". These are interesting questions, but 
what the  patient needs to know is "What strategy of treatment will give me 
the longest  and healthiest life?" Of course, it is unlikely that this will be 
the same for  everyone.
 
The question is complicated because CLL clearly falls into two camps: those  
with mutated IgVH genes, and low ZAP-70 and CD38; and those with unmutated 
IgVH  genes and high ZAP-70 and CD38. You might even say that there is a third 
group:  those with discordant markers. When comparing one treatment with another 
it is  essential that these camps are equally distributed among the 
treatments.
 
Of course, it can well be argued that our aim in treatment is cure, and  
since no-one has ever been cured by chlorambucil it ought to be abandoned.  
However, it also true that no-one has ever been cured by FCR, although we are  still 
hoping that some might be. It is also true that only the occasional person  
has been cured by autografting and that although mini-allografts have probably  
cured some, we don't know for sure, and this treatment is unavailable to many 
 and can be extremely toxic for some.
 
My view on this is that cure is a laudable aim and by all means should be  
pursued by experimental therapies with the full and informed consent of the  
patients, but there are many patients who will live a normal lifespan with their  
CLL, who will only ever need symptomatic treatment, and these can largely be  
predicted at diagnosis using the available prognostic markers. Including such 
 patients in trials with curative intent exposes them to unwarranted risks 
and  confuses the results of the trials because they are easier to treat than 
those  who really need a cure.
 
So, why does the US espouse FCR and Canada chlorambucil and prednisolone,  
and who is right? The answer is both are right and both are wrong. The US system 
 relies on the free-market. Patient power is such that if enough patients 
make a  fuss, insurers can be persuaded to change their policies. This especially 
true  of big business. If a huge company threatens to take its business 
elsewhere  because an insurer was denying FCR to one of its favored employees then 
the  insurer will cave in no matter what the evidence based medicine says. In 
the  Canadian system the government has more of a say and while they are 
influenced  by voter power, a single individual finds it much more difficult to 
influence  things.
 
In my ideal world, everyone would have prognostic markers done at  diagnosis. 
Those with good prognosis markers would only be treated when symptoms  
demanded it, and the emphasis in choosing a treatment should be "first, do no  
harm". Those with poor prognosis markers should be informed of the situation.  
There will be some who choose less intensive treatment because of age or  frailty. 
For others, especially those who are young, they should be  offered treatment 
curative intent. At the moment this involves getting  the patient into a 
state where a mini-allograft is possible, but clinical trials  may yet reveal 
better strategies. If a patient is willing to be entered into a  clinical trial 
there are several questions that need answering. Here are a  selection: "Are 
more patients cured if treatment begins when there is only a  small amount of CLL 
present, rather than waiting until it is symptomatic?"  "Would elimination of 
minimal residual disease with Campath after a remission is  achieved with 
say, FCR cure more people than a mini-allograft?" "Would adding  mitozantrone to 
FCR improve the complete remission rate?" Can anything be done  to restore the 
immune deficiency?" and there are many more possible questions.  For those 
who don't want to enter a clinical trial, then FCR is certainly an  option as 
long as the patient is aware of the pros and cons of the  treatment.