One of the things that I do is oversee clinical trials. I no longer participate by entering patients into trials but I assess whether a trial should take place on grounds of both ethics and funding, I monitor the data produced and make judgements on how the trial has been conducted and I assess the results.
I sit on the Gene Therapy Advisory Committee, which is the Ethical Review Committee for all trials conducted in the UK involving either gene therapy or stem cells.
The science involved in these studies is challenging, but it is necessary for the committee to understand it in order to assess how likely it is to succeed and how dangerous it is likely to be. Of around 160 studies that we have looked at over the past 10 years, the majority have been ineffective and did no harm, but of late there have been trials that have been both effective and very toxic. There has been a steep learning curve for us members.
Some of the more arcane elements of gene therapy are outside my field of expertise, but it does help to have a good understanding of general medicine and of the nature of cancer. One of our most pressing concerns is the quality of the patient information leaflet. We insist that it is written in terms that a lay person can understand. It is important that the information does not give an over optimistic outlook for the treatment. Remember that what is taking place is an experiment. The likelihood is that it will not benefit those taking part, though information gathered from the trial might benefit later patients. It is easy to attract patients to trials of treatments for hopeless diseases. Whatever you say to them they believe that they will be the one to benefit.
I also sit on CTAP which is the grant awarding part of the Leukemia Research Fund. It funds phase 2 trials, which look at efficacy of a new drug or combination of drugs in a small number of patients. It is not a wealthy organization so we only want to fund the best. Our concern is not to fund trials that are already taking place elsewhere, and to ensure that the maximum amount of information is obtained from the trials we do fund. The major LRF funded trial in recent years is the one that demonstrated clearly that FC is better than either single agent F or single agent chlorambucil for CLL.
In addition I sit on a number of Data Monitoring Committees and Independent Review Panels for CLL trials. Here we have to monitor how the trial is being conducted, whether unforeseen complications occur and to recommend protocol variations if they do, and to assess whether a trial should move on to the next phase.
I am in the middle of a succession of such meetings at the moment, which I why I have been a rather silent blogger.
I am, of course, sworn to silence about individual trials, but I have some observations about trials in general, which may be of interest to my readers.
First, clinical trials are essential. Without them we have only a series of anecdotes and opinions. Trials should be designed so as to find out if the hoped for conclusion is unlikely to happen. For example, if you believe that all swans are white and there are stories that black swans exist in Australia, then there is no point in only looking at swans in America or Europe. Your theory must be tested under conditions most likely to prove it wrong.
The ultimate test of a new treatment is the randomized controlled trial in which the new treatment is compared with the best of the old in a population randomly allocated to either treatment. Ideally, both patients and their medical teams should be unaware which treatment the patient is given and the success or otherwise should be assessed independently by someone unconcerned with the trial who also does not know which treatment the patient received. Sometimes it is impossible to meet these conditions. For example if a new treatment makes you sick and the old one doesn't then it is pretty plain who is having what.
It is noticeable that some people, such as homeopaths, don't buy into RCTs. All I can say is that any assertion that such an untested treatment works is likely to be spurious. I am quite willing to believe that many patients feel better when they purchase the whole package, with the time it takes, the sweet talking and the impressive procedures, but such packages are impressive placebos.
Any substance put forward as a treatment will already have demonstrated activity in a test tube and should have been given to at least two species of animals to demonstrate its safety. The first type of trial that is done is a phase 1 trial to determine a safe dose. Such trials are often done of normal volunteers, but they can go disastrously wrong as happened at Northwick Park. Six young men reacted disastrously to minute doses of an antibody the first time it was administered at a much lower dose than had apparently been safe in monkeys. Who would volunteer to take part in such a study? Many young men earn a sizable living from such studies, getting up to $5000 a time.
But most phase 1 studies are pretty safe. Cancer drugs are almost always tested in cancer patients, often in patients who have reached the end of the line, having exhausted all the licensed drugs. They are hoping for another Gleevec, for it was in a phase 1 trial that Gleevec first showed its superiority in CML. The idea of a phase 1 trial is to find the highest dose that is safe. Usually cohorts of three patients are treated with a single escalating dose. Eventually a dose will be found that causes toxicity and the last dose below this is chosen.
1. 2. 3. 5. 8. 13. 21. 34 ... is known as the Fibonacci sequence and the rate of escalation is a modified Fibonacci sequence.
A phase 2 study is looking to see whether the chosen dose has any activity. Fourteen patients need to be treated without benefit before concluding that the drug has no effect. After that more may be treated to determine exactly how great the effect is. However, around 40 patients is usually enough. Randomized phase 2 studies enable us to compare an immediate effect with the standard treatment. For example, drug A might cause 40% remissions against only 25% with the standard therapy. This doesn't tell us that drug A is better. Remission rate often correlates with survival, but not always. If drug A puts 40% into remission but kills 60% it is not such a good treatment as drug B which only puts 25% into remission yet only kills 5%, the remaining 70% living for more than 2 years.
What we are usually looking for is length of overall survival, and if that is the same for both treatments then the quality of life during that period of survival comes into it. To determine whether a given treatment is better than a competitor, then we need to do a randomized phase 3 trial. The endpoint is overall survival and the number of patients required is determined by how much better than standard treatment the new treatment is suspected of being. A small difference requires a big trial; a big difference only a small trial. When improvements in outcomes are only nibbling at the problem then very large trials are needed.
For diseases like CLL where current treatments are associated with very long survivals in most patients you have a problem. You can either pick out some characteristics associated with very poor survival (like del 17p) and do your trials on this group, or you can find a characteristic associated with very long survival (like MRD negativity) and use that as a surrogate end point.
One of the problems with CLL is that as time passes we know more about the disease, so that ongoing trials become superfluous. Take the definition of a complete remission. Only about 15 years ago, all you needed to have a complete remission was no symptoms, a normal blood count, no lymph nodes to feel in armpits groins or neck, no liver or spleen to feel, and a bone marrow with fewer than 30% small lymphocytes in it. Today that is not sufficient. CT scanning has shown that enlarged lymph nodes may be present in the abdomen in many patients with what would previously have been called a CR. The spleen has to be three times the normal size before you can feel it and livers can be felt even though they may be a normal size. You can't make livers and spleens disappear completely so who knows whether a liver or spleen is of normal size or slightly enlarged? CT scans are so sensitive that a normal lymph node may be seen and called as pathological.
To get round this problem, a new standard is emerging. Tiny numbers of CLL cells can be detected in the blood using idiotypic PCR or 4-color flow cytometry. Increasingly, trials are being designed with negative minimal residual disease as an endpoint, and in my opinion this is to be encouraged.
Finally, when a patient volunteers to take part in a clinical trial, the doctor conducting that trial has been given a great gift and it behoves him to conduct the trial properly. I have seen too many occasions when the recording of observations is frankly slapdash. I think that some centers enter patients into trials for pecuniary reasons or just to get their names on a paper, without any curiosity about the outcomes. Improperly conducted trials are unethical and should not be allowed. When a pharmaceutical company finds that a particular center does not comply with the protocol the center should be excluded from future trials. For a patient to quit a trial midway is also an affront to all the other patients involved. I understand that the toxicity may enforce abandonment, but there will be follow-up blood tests and examinations that ought to be done. For a pharmaceutical company not to publish the results of a trial (even if it is offensive to their shareholders) is an insult to the professionals and the patients who took part.