There is nothing like a story that tugs on the heart strings and forces reconsideration of an issue. Take the very sad tale of two close cousins from California who both developed melanoma - Tom McLaughlin landed in the experimental arm of a clinical trial for a Roche drug, PLX4032, while Brandon Ryan was chosen by computer lottery for the placebo. Their plight and the implications were detailed movingly inThe New York Times.
The central issue raised is whether a controlled trial for extending life is ethical. In this instance, PLX4032 was shown to shrink tumors in some patients, but only for a limited time, and the key question was whether those patients would live longer. The standard chemotherapy used in melanoma, dacarbazine, slowed tumor growth in 15 percent of patients for an average of two months, while the Roche drug halted tumor growth in 81 percent of patients for an average of eight months. As the paper notes, some meds may be so much more effective than existing drugs that some docs say putting patients into control groups, and delaying access to thousands of others, is unfair. And many docs believed that if the Roche drug provided relief by shrinking tumors, patient lives would improve.
“With chemotherapy, you’re subjecting patients to a toxic treatment, and the response rates are much lower, so it’s important to answer ‘Are you really helping the patient?’ ” Charles Sawyers, who chairs human oncology at Sloan-Kettering, tells the Times. “But with these drugs that have minimal side effects and dramatic response rates, where we understand the biology, I wonder, why do we have to be so rigorous? This could be one of those defining cases that says, ‘Look, our system has to change.’ ”
Here was the rub: Roche was urged last year to seek accelerated approval. The Times writes, however, that getting patients to join a trial for a drug already on the market was unlikely, since the odds were only 50-50 that a patient would receive the med and patients were already clamoring for the Roche treatment. And without the trial, obtaining "definitive" effectiveness data was also unlikely. Yet the trial would cost $100 million and FDA approval may have to wait two years, which some docs felt was a waste when the important issue was prolonging remissions, the paper adds.
Some docs suggested a combo treatment involving rival meds may be more effective, but Roche refused to give permission for testing. The drugmaker "feared that might lead to approval for only a narrow group of the sickest patients. The surest way to get the FDA’s endorsement for a broader market was a controlled trial. And with its competitors rushing to get similar drugs to market, the findings of such a trial might give Roche an advantage in marketing its version as the only one proven to prolong survival," the paper writes.
The FDA's Richard Pazdur, who heads the Office of Oncologic Drugs, calls the situation "unprecendented," and adds that "regulatory flexibilty" may be in order, according to the paper. Meanwhile, Ryan passed away. “What gives them the right to play God?” Ryan tells the Times. “It doesn’t make sense to say, ‘We want you for a statistic’ instead of giving them a chance at life.” But what do you think?
Are controlled trials for extending life the best way to evaluate cancer drugs?
- Yes (69%, 87 Votes)
- No (31%, 40 Votes)
Total Voters: 127
32 Comments
This is a hard one but let's think about the closing snetence:
“What gives them the right to play God?” Ryan tells the Times. “It doesn’t make sense to say, ‘We want you for a statistic’ instead of giving them a chance at life.”
Without the statistics, there are crys of "rushing to Market" and "money first." The sad fact is that we need good statistics and they are "hard to come by." I wonder what others here will think?
It may not be moral to force a control arm, but it is ethical for the simple reason it would be unethical to allow a drug to be used without a reasonable degree of certainty that the drug is safe and effective for its intended use. The criteria of "Reasonable" will vary, but there is no better way around the science. Sad he died? Yes. Moral that he was forced into a control arm? Maybe not. Ethical that he was? Definitely.
And let's not oversee he was receiving standard of care in a very controlled clinical setting, was not a placebo, or simply left untreated as many without financial support surely are. The old drugs may not be any good, but they do constitute a reasonable well known and understood entity. If the trials are not done properly, then there is no hope of improvement in the clinical arena.
Controlled trials may be necessary, but compassion is better and more humane. . I have heard of compassionate access--where patients who have run out of other options are allowed access to the trial drug (not the control drug), with certain caveats about what risks they would be willing to assume. . We need this option and flexibility, especially as treatments become more intelligent, based upon underlying biological markers, and such.
Ethics are always a matter of balance between difficult choices, leading one to pick the better of the choices. Just because one of the choices may be an "unethical" one, does not make the process less ethical.
In this case, there is no assurance that the product works, so there is no a priori knowledge that would make not giving the active ingredient an unethical move. Likewise, the study benefits a whole population vs. some individuals alone.
Feel really bad for those people who have been in control arms of positive studies...but there are many more in society who have benefited from those studies.
Clinical trials design for cancer drugs and for drugs for serious illness in general is always a hard task and there are many ethical considerations behind statistics and endpoints. There is no one right answer, it's definately case by case. On one hand we need EBM and statistics to measure clinical effect - it's inappropriate to put on the market untested drugs - drug would be administered to larger population of patients. Clinical trials, EBM and preclinical programs etc. resulted form many tragedies (mainly related to safety) in the past. It doesn't mean clinical trials assure safety and efficacy (we have recent examples of failures). On the other hand there is ongoing discussion - there are always ethical considerations behind simple statistical significance and clinical significance in such diseases, especially for drugs where overall survival benefit is i.e. of 3-4 months (still statistically significant). So probably we have to combine both approaches.
I don't fault desperate parents for not being able to split these hairs, but:
A clinical trial isn't supposed to be about getting a cure or extended life for an _individual person_. It's to find out how a drug or treatment performs on a particular _population_.
There are too many hidden assumptions in the phrase "Your best chance is to be in a clinical trial". Best chance of what, exactly? "Of possibly getting a new treatment" -- that's not the same as "of definitely getting a new treatment" or "of being cured"....
An additional hair in need of splitting is the difference between patient care -- the role of ethical physicians -- and investor care -- the genuine objective of pharmaceutical companies. There may be times when these two things overlap, but to _expect_ or DEMAND that they are the same thing is not realistic. Blurring these concepts seems to have been the goal of many pharma branding campaigns, but we can see here a variety of adverse events related to such treatments.
Just to get back to basics, This was a Phase I Safety/Pharmakokinetics trial (see link below). Survival was not even a secondary endpoint. The survival benefit seen in this small study therefore could not be used for any type of regulatory decision making. As suggested above, it may just be a statistical artifact or an epiphenomenon. Let Phase II/III continue before making any decisions. A larger study would allow placebo-treated patients to be automatically rolled over into the acive treatment arm if evidence of efficacy is demonstrated.
http://www.clinicaltrials.gov/ct2/show/NCT00405587?term=plx+4032&rank=1
@Arlene - compassionate access for a drug in advanced trials and a high degree of support for safety and efficacy is one thing but demanding access to unproven drugs on "compassionate" grounds can be a recipe for disaster.
Our society is on an intellectual "race to the bottom", and that article in the NYT is just a sad, sad example. A drug still in Phase I, which may very likely pan out to kill people faster than it saves them, is NO ONE's "chance at life". The writer of this article should be returned to the fashion desk at the NYT, and these two young men should never have been enrolled in a Phase I drug trial. Everyone in our office this morning fears that we will never see useful clinical data on Oncology medications if this trend continues.
I actually hope that this drug works out. However, just to put in perspective, the whole therapeutic approach of targeting single gene mutations in oncology has very limited applicability. The reality is that in most cancers multiple gene mutuations can be detected, and these are not necessarily static. By analogy, targeted therapy is often like trying to put out a five alarm fire with a cup of water.
I agree with Dismayed in that this type of reporting is about as helpful as a chocolate teapot. This type of reporting - and I especially hate the "cure for cancer" stories that are about a preclinical thing that is years away - serves no useful purpose to society.
This said, are we not also getting to a stage where the developmental processes for approval and science are too contradictory to co-exist? Frequently companies are having increasing difficulty in recruiting participants into trials and the numbers of participants required are out of proportion to the size of the disease population that is accessible. The problem lies with the use of traditional statistics that require a certain sample in order to be "representative". It seems to me that the time is upon us when the use of probabilities and Bayesian methods (which require significantly smaller samples but are still reliable) must be considered for some, or perhaps all, products.
In a TheScientist.com article, "Crowdsourcing Drug Discovery," there was a comment that for the last 50 years, randomized controlled trials have been the unquestioned gold standard, when in fact they have become a fiercely defended relic of our ignorance in 1962 when Congress empowered the FDA to begin regulating efficacy.
At that time, it was a "best we could do" solution - but now? They take too long, cost too much, are fraught with unsolveable ethical problems that patients and many physicians dislike, and cannot ask the patient-specific molecular questions we now know need to be asked and answered.
Yet, most clinical trialists and the FDA cling to these crude, simplistic tools like an irrational safety blanket. If we can't reach agreement that clinical methodologies must adapt to new knowledge of the biology of disease, and that the way drug development is regulated must rapidly adapt in much the same way, then our ability to accelerate advances in medicine will remain stagnant.
A key point is, the system should be patient-centric. It has to be something patients will not only tolerate, or enter under duress, but rather a system that makes sense to them personally - even when they are not yet facing a serious or terminal condition. If real patients are left out of the process of change, we will likely end up in the wrong place again.
The problem is not with using the prospective, randomized trial as a research instrument, the problem comes from applying this time and resource-consuming instrument to address hypotheses of trivial importance: do cancers prefer Coke or Pepsi?
There are no standard treatments for advanced melanoma that have any reasonable success. As an oncologist, I know that treatment with existing, approved drugs leaves much to be desired. Select any combination: DTIC, Interferon, Cisplatin, Vinblastine, DTIC and Tamoxifen and you have a toxic treatment with only a small number of individuals who are helped.
My feeling is that the Plx4032 should be approved for select melanoma patients (those with BRAF mutations) is PLX4032 demonstrates a reasonable number of tumor regressions, say regressions in over 30 percent of patients where the regressions last at least 3 months, if it turns out that the drug is well tolerated. You could quibble with the numbers.... maybe you want 50 percent for at least 4 months. But if the drug is non toxic than this information is sufficient for me. I don't need a placebo arm for Melanoma.
Now, if you compared DTIC arm to DTIC plus PLX4032 I could see such a study. There is no placebo arm in this sort of study.
I was recently approached to consider a study of a JAK2 mutation inhibitor for patients with severe myelofibrosis. In this study there would be a placebo arm. I turned the study down. Do we really need to do repeated MRI's of a patient with myelofibrosis, a large spleen and a placebo treatment? Doesn't make sense.
There may be no "standard" treatment for advanced melanoma that have any reasonable success, however, there are existing approved treatments that have shown reasonable success in cell culture assays (Melphalan, Interleukin-2, ImuVert [Serratia marcescens microscomes], Interferon, Temsirolimus + Bevacizumab).
Cancer dynamics are not linear. Cancer biology does not conform to the dictates of molecular biologists. Genotype does not equal phenotype. Using genomic signatures to predict response is like saying that Dr. Seuss and Shakespeare are truly the same because they use the same words.
The building blocks of human biology are carefully construed into the complexities that we recognize as human beings. However appealing gene profiling may appear to those engaged in this field, it will be years, perhaps decades, before these profiles can approximate the vargaries of human cancer, as the group from Duke and many others have seen.
Functional profiling analyses, which measure biological signals rather than DNA indicators, will continue to provide clinically validated information and play an important role in cancer drug selection. Primary tumors are heterogeneous and have a genomic signature unique to every patient.
Patients would certainly have a better chance of success had their cancer been "chemo-sensitive" rather than "chemo-resistant," where it is more apparent that chemotherapy improves the survival of patients, and where identifying the most effective chemotherapy would be more likely to improve survival.
A change is needed.
When if comes to molecularly targeted drugs, we have the benefit of knowing with pre-testing which patients might actually fair better.
I realize that this was a Phase I trial. How could we possibly know that this was a "block-buster" - there wasn't enough evidence... However, can I just say that with the way we design these drugs now, we have a better idea of what might work versus even a decade ago....
If we have reasonable evidence that something might work, I am for allowing people whose options have all been played out to be given the drug under close supervision with all the caveats fully exposed and the drug companies should absolutely monitor and capture all the data. One would hope that this data is taken into consideration at some point.
My honest opinion is we have to find a better mechanism to get these drugs to the patients who need them desperately without it tainting the prospects of the drug going through a rigorous approval clinical trial process.
Importantly, I think we all need to think out of the box on this one.
It is time for a change - let's do the right thing.
Here's how I would paint the scenario based on 28 years of running pharmaceutical clinical trials: Roche pushed this story out because they desperately need some good PR in the wake of the recent hits they have taken on Avastin in both breast and late stage colon cancer. Absent these major disappointments I don't think that we would have seen Roche plant this overhyped story with an uncritical NY Times reporter. If this study had any credibility you would read about it in the "Science Times" section of the NYT, which is published every Tuesday. My guess is that The Science Times editorial staff considered and rejected this piece.
I hope that none of the people who posted a comment will ever need an experimental drug or be in a desperate position.But just once, I'd like to hear from a patient/consumer person who has gone through this tough and difficult ordeal.
I will say what is a tough ordeal: the nearly 40 year process that the FDA has been going through trying to remove, over the objections of some patients and doctors, those products that were approved prior to 1962 that were subsequently found to be ineffective. Prior to 1962 only safety was required; after 1962, both efficacy and safety were required for approval. Post-1962, many such drugs were subject to DESI (Drug Efficacy Study Implementation), which reclassified pre-1962 drugs as effective, ineffective, or requiring further study. It took the better part of the last half of the 20th century to finally get these drugs off the market.
Controlled clinical trials are the bedrock of the drug approval process. A fair reading of FDA history will support that proposition.
I participated in one clinical trial. It did not involve a life or death issue. I would never participate in any trial that involved a potentially life threatening or life ending issue. I don't believe it is ethical for the FDA and pharmaceutical companies to do so.
Something is not right with this picture. Industry Insider yesterday said “This was a Phase I Safety/Pharmacokinetics trial” and s/he gave this link http://www.clinicaltrials.gov/ct2/show/NCT00405587?term=plx+4032&rank=1
The description of that study at ClinicalTrials.gov says nothing about a control group that receives only standard treatment.
My guess is that the study involving the 2 cousins featured in the NYT story is this one: http://www.clinicaltrials.gov/ct2/show?term=plx+4032&rank=3 Here is the description: This randomized, open-label study will evaluate the efficacy, safety and tolerability of RO5185426 as compared to dacarbazine in previously untreated patients with metastatic melanoma. Patients will be randomized to receive either RO5185426 [RG7204; PLEXXIKON: PLX4032] 960 mg orally twice daily or dacarbazine 1000mg/m2 intravenously every 3 weeks. Anticipated time on study treatment is until disease progression or unacceptable toxicity. Target sample size is 500-1000 individuals.
We might well ask why up to 1000 patients are needed to establish efficacy. Alternative trial designs such as a matched pair sequential trial would require fewer than 100 patients to inform us whether Treatment A or Treatment B is superior or whether there is little to choose between them.
Bernard, here is link to the published data. I believe that the NYT article quotes the 32 patient Phase I study because it quotes the same 81% response rate in both places. I haven't seen any published data from Phase II/III. Also remember that clinicaltrials.gov is not always 100% in their study descriptions. Let us know if you find anything further.
http://knol.google.com/k/krishan-maggon/plx-4032-plexxikon-roche-melanoma-review/3fy5eowy8suq3/135#
I checked the new link given by Industry Insider and I do not see a reason to change my earlier comment. The study to which II linked originally,NCT00405587,is a non-randomized, uncontrolled study of safety and pharmacokinetics. So there is no way one of the 2 cousins would be denied entry based on randomization, as described in the NYT story. A short term response rate of 81% was found in NCT00405587, and that number was mentioned in the NYT, but that is not the study the 2 cousins were involved with.
Thanks, Dr. Carroll. I agree with your larger point that some type of sequential design should obviate the need for such a larger sample size. Furthermore, in the excellent paper by David Schrama et al in 2008 Journal of Carcinogenesis (see link below), simply the presence or absence of the BRAF gene mutation is not predictive of malignant transformation of nevi and other premalignant melanotic conditions. The right kind of activating gene sequences are also required for malignant transformation. Hence there is a need for a larger scale definitive randomized trial, but as you indicate, an n=1000 seems like an overpowered study if the clinical efficacy is as striking as the early data seem to suggest.
http://www.carcinogenesis.com/article.asp?issn=1477-3163;year=2008;volume=7;issue=1;spage=1;epage=1;aulast=Schrama
Yes. The sequential trials design is under-utilized. Forty years ago there was happy talk that the amino acid precursor of serotonin, called l-tryptophan, was as effective as ECT in treating depression. On its face this was an improbable claim. With colleagues in Melbourne I ran a sequential trial and it took only 12 pairs of patients to establish that ECT was clearly the superior treatment.
A lot of the moral indignation about drugs that have run into trouble post-approval has focused on pivotal trials having too few participants. Industry has responded by overpowering trials.
If a product were approved based upon a couple of 100-patient studies & subsequently had problems that led to withdrawal, the public outrage would probably leave a crater where FDA used to be. The study designs may be rational, but the public often isn't.
This article, like many others, has missed the point of what is an eithical trial. To be ethical the trial has to start off with equipoise for all treatment arms. Without this foundation, then the ethics of patient treatment in one arm v the other questionable.
It is always sad when for reporting this kind of situation (individual patients) is brought up and raises the supposedly ethical questions without understanding the underlying ethics. It is often without these larger studies the true safety profile of a new therapy can be fully elucidated, and what initally appeard to be a drug that offerered hope is removed to the list of disappointing failures. This is not to negate the issue that clincal trial design has to be further improved, but that it needs to be considered in the fulcrum of scientific rigor and debate on several fronts, not least the ethical underpinning of the whole process from ethicists.
Salient point makes a salient point. A 1000 patient melanoma trial will probably result in a $20-25 million cost overrun vs a 100 patient sequential design, which of course will result in an unnecessary increase in the price of the drug. My first boss in Pharma Clinical Development once told me that "the only thing that I care about is if the P-value is less than 0.05". Accordingly he would overpower Phase III trials at 90 per cent or greater, rather than the standard 80%, which our Phase II data predicted would be adequate to show that a clinically meaningful drug-placebo difference would be statistically significant with alpha set at 0.05, and a power of 80%.
As you know from statistics, the power curve is not linear, and a 90% powered trial usually costs considerably more than an 80% trial.
I used to have a friend who would take out enough life insurance for a family of four, although he was unmarried. He was so worried about being underinsured that he wound up being way overinsured.
Salient Point has put his finger on the right issue. There is a difference between clinical science and regulatory procedures. The two activities overlap but they are not coterminous. The risk-benefit tradeoff also should influence how regulators make decisions. SP is right in respect of a drug like Vioxx that did not meet a critical medical need, and which patients might take for a long time. Long term cardiovascular risk is a salient issue in that context, and large samples are needed for accurate estimates of such risk. For a drug like RO5185426 in treatment of metastatic melanoma, however, greater risk including unknown risk is acceptable, and approval could be defended on the basis of smaller studies.
By the way, when penicillin and streptomycin came into general use in the 1940s and 1950s, they were so obviously efficacious for streptococcal pneumonia and tuberculosis that formal controlled trials were not conducted. Outcomes were compared mainly against historical fatality rates. But if a new antibiotic is intended nowadays for a less critical disorder like H. pylori gastric infection, say, then a controlled trial would be required by regulators.
It’s all about the context, and one size or shape of trial does not fit all cases.
Good point. For me, here is the intersection of clinical science and regulatory procedures: I believe that it would be appropriate for an FDA clinician or statistician at an End-of Phase II meeting to advise the sponsor that a smaller, well-powered Phase III trial would expose fewer patients to the risk of an unproven treatment. However, having sat through many such meetings, I know from experience that that these types of questions are never asked. Uusually it's the other way around; FDA asks for larger trials, not smaller ones.
Good discussion. My impression has been that the size and design of phase III trials is negotiated beforehand between FDA and company (to give best possibility of avoiding wasted effort). Is that correct.
Yes. A company will have an open dialogue with the FDA to determine what kind of data is potentially needed in Ph3 for FDA approval. To be fair to the FDA, they ask for anything from simple PK in some hormone replacement situations, to enormous 10k plus studies in the cardiac space where the additional benefit from the new drug is likely to be small.
I am sitting here wishing and waiting for approval of PLX4032 by the FDA as I am braf positive and I have metastatic melanoma. Those of you would deny me the chance of longer life, even of just a few months, should come day be in my situation, so that you can have the ethical numbers you desire. We are not talking about the latest drug for depression or acid reflux or even some other cancers. We are talking about my life. If I am informed about the consequences of taking the drug, and I am dying and have endured all the other limited options available, why would you deny me this?