About this Author
DBL%20Hendrix%20small.png College chemistry, 1983

Derek Lowe The 2002 Model

Dbl%20new%20portrait%20B%26W.png After 10 years of blogging. . .

Derek Lowe, an Arkansan by birth, got his BA from Hendrix College and his PhD in organic chemistry from Duke before spending time in Germany on a Humboldt Fellowship on his post-doc. He's worked for several major pharmaceutical companies since 1989 on drug discovery projects against schizophrenia, Alzheimer's, diabetes, osteoporosis and other diseases. To contact Derek email him directly: Twitter: Dereklowe

Chemistry and Drug Data: Drugbank
Chempedia Lab
Synthetic Pages
Organic Chemistry Portal
Not Voodoo

Chemistry and Pharma Blogs:
Org Prep Daily
The Haystack
A New Merck, Reviewed
Liberal Arts Chemistry
Electron Pusher
All Things Metathesis
C&E News Blogs
Chemiotics II
Chemical Space
Noel O'Blog
In Vivo Blog
Terra Sigilatta
BBSRC/Douglas Kell
Realizations in Biostatistics
ChemSpider Blog
Organic Chem - Education & Industry
Pharma Strategy Blog
No Name No Slogan
Practical Fragments
The Curious Wavefunction
Natural Product Man
Fragment Literature
Chemistry World Blog
Synthetic Nature
Chemistry Blog
Synthesizing Ideas
Eye on FDA
Chemical Forums
Symyx Blog
Sceptical Chymist
Lamentations on Chemistry
Computational Organic Chemistry
Mining Drugs
Henry Rzepa

Science Blogs and News:
Bad Science
The Loom
Uncertain Principles
Fierce Biotech
Blogs for Industry
Omics! Omics!
Young Female Scientist
Notional Slurry
Nobel Intent
SciTech Daily
Science Blog
Gene Expression (I)
Gene Expression (II)
Adventures in Ethics and Science
Transterrestrial Musings
Slashdot Science
Cosmic Variance
Biology News Net

Medical Blogs
DB's Medical Rants
Science-Based Medicine
Respectful Insolence
Diabetes Mine

Economics and Business
Marginal Revolution
The Volokh Conspiracy
Knowledge Problem

Politics / Current Events
Virginia Postrel
Belmont Club
Mickey Kaus

Belles Lettres
Uncouth Reflections
Arts and Letters Daily
In the Pipeline: Don't miss Derek Lowe's excellent commentary on drug discovery and the pharma industry in general at In the Pipeline

In the Pipeline

« OCD Linked to the Immune System? | Main | Ipilimumab (And Progress Against Cancer) »

June 7, 2010

Again, What's It Worth to You?

Email This Entry

Posted by Derek

Let's open up a painful subject here. This is prompted, partly, by the news the other day about Erbulin. The main reason I posted about that compound was because of its chemical complexity and total-synthesis heritage, both of which are unusual. But it's an oncology drug. As such, it looks like an awful lot of other drugs in that space.

And that's not good. Because we should face up to the fact that most of the newer anticancer compounds aren't so good, not in any absolute sense. (Most of the old ones, too, but those are rather cheaper, aren't they?) Too often, what we're looking at is an extension of a few weeks or months of life - life as a terminal cancer patient, mind you, but life nonetheless. The price you put on that will vary according to your circumstances, but in many cases we're asking a lot. Is it worth it?

Increasingly, that's not a question that's going to be answered by the patient alone, but by some combination of the patient and his or her insurance company. In other cases, it will have been answered well before by a country's national health service, when it did a cost/benefit analysis of the drug and decided whether it would even be included in a national formulary. That's the whole point of the UK NICE, and although their execution has not been trouble-free, the idea behind it is not going to go away.

Nor should it. Now, I think people should be able to spend their own money on what they want to spend it on. True, cancer patients are known for spending some of it, out of sheer desperation, on bizarre and worthless stuff. Taking advantage of these people by selling them Wonder Water or Miracle Mixture is a crime, as far as I'm concerned, and should be prosecuted. But I'm not advocating force to keep people from exercising their choice to buy the stuff, if it's out there, although I'd certainly like to talk them out of doing that. I'll reserve the force for the ones selling it, so that the crap is not out there for purchase in the first place.

But when it's other people's money being spent - an insurance company's, or tax money - those others should get a say. And here's where things get messy. Because while there's a difference between Wonder Water and the latest angiogenesis inhibitor or cell-cycle interrupter, it's not a meaningful a difference as anyone would like. True, one is likely to do nothing, and the other is likely to do something. But when "something" is "keeping you from dying in November, in order that you should die in March", well. . .you'd want something more, wouldn't you?

Let me say here that it's not for lack of trying. We in the business keep throwing our best punches in oncology. Here, here's a target that makes perfect sense - this thing should kill a cancer cell. Shouldn't it? I mean, cutting off the blood supply to a solid tumor is a good idea. Messing up mitosis, re-establishing programmed cell death: good ideas. But they just don't work as well as we'd hope that they would. We're not there yet.

And so we get these add-a-few-months-of-life drugs, because in most cases, that's all we're capable of delivering at the moment. But we're asking a lot of money for these things, and increasingly, the people who are really paying for them are asking whether anyone's getting a good deal.

I wrote about this situation a few years ago on this site, and I have to say, not much has changed - other than the pricing pressure, of course. I'll have some more to say about this issue this week, but I wanted to start people thinking - about where we are, and whether there are any ways out.

Comments (51) + TrackBacks (0) | Category: Cancer | Drug Prices


1. Blah on June 7, 2010 11:09 AM writes...

You can't be serious. I understand that one a cost-benefit analysis, most treatment of oncology patients would seem rather irrational (based on public funding) but you can not be serious. Human lives are human lives - priceless. They cannot fit into a cost-benefit analysis. For some people, a few months is a huge deal. Family members will go broke trying to fund these few months of their loved ones - for a reason! Granted, the extended time of some patients are horribly painful and they would probably agree that it wasn't worth it but for some it is an absolute blessing.

On a different view, for a lot of the drugs the benefit is small. But if we stop paying for it, the companies that make those drugs will disappear and we might miss the next big drug that really works well. There are many many treatments that virtually cure cancer patients. We cannot forget that! Oh man, there is just so much wrong with this that I better stop here. I'll tap somebody in to keep these arguments going...

Permalink to Comment

2. Vader on June 7, 2010 11:26 AM writes...

For the patient, the benefit of a few more months to wind things up seems priceless. Via insurance, the cost is diffused over the rest of us. Whenever there's a concentrated benefit with a very diffuse cost, the social and political pressure will strongly favor paying the cost, whatever the absolute cost to benefit tradeoff. Perhaps that's as it should be.

I think it's important to compare apples to apples, though. Giving someone a few more precious months at the cost of higher insurance premiums for the rest of us is apples to oranges. Giving someone a few more precious months at the cost of those resources not being available to save a kid dying of river fever in a Third World country is apples to apples.

Permalink to Comment

3. Evorich on June 7, 2010 11:39 AM writes...

Along with the fact that many cancer drugs are much better than "a few extra months" - many drugs are showing genetic profile preferences. This is something which hospitals have realised and work with a lot through their in-house sequencing centers. This is now something that pharma companies can begin to design towards - a slant which will benefit both the industry and patients.

There are new cancer targets being discovered all the time. People in my team are working on 5 oncology targets right now which have not previously been targeted.

NICE is good in theory but I'd rather money was saved by cutting the huge number of NHS management than by cutting people's possibility of survival.

And this is absolutely NOT NOT NOT the same as selling magic water! To me there is an infinite difference between a 1% chance of success and a 0% chance of success. One has possibilities - the other does not.

Permalink to Comment

4. Jeffrey on June 7, 2010 11:59 AM writes...

This reminds me of one of my favorite pet peeves: the contrast betweent the vast amount of resources devoted to adding a few months of (low-quality) life, and the minimal resources devoted to better diagnostics.

Catch a cancer (nearly any cancer) in stage 1, and it isn't a big deal. Catch it in stage 4, and its time to wind up your affairs.

Common sense indicates we should try as hard as possible to catch cancers as early as possible.

The state of the art is so primitive that one of the better detection methods is to feel for lumps. We don't always know which growths have to be worried about, and which don't. We can't apply any fancy proteomics technology to measure expression levels in your blood. We can't do fancy diagnostic scans and get actionable results.

What would happen if someone spent $1bn to address these problems?

The reason why no one does is fairly obvious: economics. For someone who has just a few months to live, we get "human life is so precious.. spend whatever it takes." But this attitude doesn't carry over to providing to scanning masses of folks who feel perfectly fine. Insurance companies will (often) tens or hundreds of thousands to extend life; they won't pay nearly this amount to develop tests that avoid the disease in the first place. Nor will patients, nor the government.

So we keep shoveling money at these extraordinarily difficult problems, and spend very little money or resources on low-hanging fruit that could avoid a lot of needless suffering and death.

Permalink to Comment

5. Betsy on June 7, 2010 12:18 PM writes...

Most of the efficacy data you see for new oncology compounds is for a single agent. Once a drug is approved, it's going to be used in combination with other therapies, and together these might extend the patient's life even further. Cancer will never be cured with a single agent.

Jeffrey, how do you propose "scanning masses of folks" for every single type of cancer? Every cancer is different, and even a single type of cancer can have very different genetic signatures in each patient. Add to that the difficulty of detecting these signatures in early cancers in a high throughput format (e.g. blood test), and you've got a bigger problem than economics.

Permalink to Comment

6. partial agonist on June 7, 2010 12:34 PM writes...

While I agree that the numbers are sometimes very discouraging, one major shortcoming of the analysis is that all forms of cancer are described by "tissue where the first tumor was noticed" and are incredibly heterogeneous. As a result, all "tumors of tissue type X" are inherently non-responsive as a large group to a single therapy or maybe even a single type of therapy.

As an example, let's say antiinfective agents were judged the same way mechanism-specific antitumor agents are. What would be the analysis if you gave Zithromax to 100 patients with "infections" of all types, including gram(+) bacterial infections, gram(-) bacterial infections, prion infections, viral infections of 50 varities, fungal infections, etc. How would it fare? If instead you only use it for certain gram-negative infections of the inner ear and throat, it is of course a great drug.

One person's melanoma may have as little in common with another person's melanoma as one person's cough has in common with another person's cough. Many of the smart drugs have great potential to treat a certain cancer phenotype. The challenge is to identify that phenotype quickly by simple in-office tests or quick turn-around lab results. It also requires fundamental large shifts in mindset of how cancers are viewed and treated.

Permalink to Comment

7. qetzal on June 7, 2010 12:35 PM writes...

Blah writes:

Human lives are human lives - priceless. They cannot fit into a cost-benefit analysis.

Unfortunately, that's absolutely wrong. Human lives absolutely do fit into cost-benefit analyses. The question is whether that fact is explicitly acknowledged (a la Derek) or denied.

Resources are finite. Time and money spent to keep one person alive for 4 extra months is time and money not spent on anything else - such as keeping multiple kids from dying of river fever (to use Vader's example), or providing hundreds or thousands of undernourished kids a decent diet.

Such trade-offs are unavoidable. Claiming that human lives are somehow exempt is understandable, but ultimately foolish.

Permalink to Comment

8. Nat on June 7, 2010 12:36 PM writes...

Human lives are human lives - priceless. They cannot fit into a cost-benefit analysis.

Like it or not, our government makes cost-benefit analyses like this all the time. When evaluating the cost of some new regulation, e.g. setting the maximum permissible trace amount of some known toxin in exhaust, they have to compare the potential number of lives saved by the new regulation, versus the economic costs of the new regulation. I realize that the idea of unelected bureaucrats crunching numbers to determine the value of human life sounds chilling, but the alternative is economic disaster. And as Vader points out, the resources aren't available for purposes that might do more good for someone else.

I don't pretend to know the answers to any of this (although I think Jeffrey comes close), but it's naive to pretend that the conflict doesn't exist.

Permalink to Comment

9. PharmaHeretic on June 7, 2010 12:40 PM writes...

I also think that developing more sensitive AND accurate methods for early diagnosis (before metastasis) is a much better use of money than the questionable advancements that pass for drug discovery in that field.

Permalink to Comment

10. David P on June 7, 2010 12:41 PM writes...

When someone has lived a good life and near the end of that life has been diagnosed with a terminal cancer, I think a lot of resources are wasted on keeping that person alive for a few more weeks or months at a very low quality of life. We should not look to prolong life when there is little benefit and much suffering.

That said, that is the decision of the person themselves, as well as their family and (it looks like) those that will foot the bill. It seems to me there is some kind of middle ground to be found, where the family can get matters in order and say goodbye but not spend resources that could be better used elsewhere in keeping someone alive for a few more days or weeks.

Permalink to Comment

11. HelicalZz on June 7, 2010 12:44 PM writes...

This post surprised me. I wouldn't expect you to fall victim to headline biases like that. Most of the 'only a few additional months' commentaries fail to appreciate that this apparently meager gain is more due to the populations evaluated in the most recent trials than a measure of a drugs full potential. For both practical and ethical reasons, new cancer treatments aren't evaluated in 'just diagnosed patients', but in populations that have little time left to begin with. Only after getting approvals in these last resort populations can companies reasonably consider going after proofs at earlier stages.

It is true that the benefits of most all newer drugs still haven't amounted to anything like curing or adding decades, and that reimbursement considerations for these treatments is a large issue. Still, I expected from this source at least some comment related to the consideration that one reason these 'improvements' seem so meager is due to how they are initially evaluated.

I'd also point out that limiting the customer base rarely makes anything cheaper -- so unless we are willing to provide as a group some sunk costs into seeing how some of these drugs work in the marketplace, we'd better be accepting of the status quo, because we won't be getting anything new.


Permalink to Comment

12. john on June 7, 2010 12:49 PM writes...

An interesting aside is that estimates by the National Academies report was between 65 and 70 percent of cancer being preventable or environmental in it's cause. Perhaps we should put a little more than 5% of our NCI budget towards prevention and prevention research.

Permalink to Comment

13. John on June 7, 2010 12:59 PM writes...

In most of these cases the drug is of a fairly simple structure, and will become very cheap once the R&D expenses are recovered and the drug goes generic. So it is not so much a question of "$20,000 per month of life saved", as it is "$20,000 per month of life saved for the first 10 years, then $200 per month of life saved in perpetuity". Only the manufacturing costs are perpetual.

Permalink to Comment

14. milkshake on June 7, 2010 1:11 PM writes...

New kinase anticancer drugs are mixed bag (and priced up as much as the market can bear) but for certain forms of leukaemia treatments with Gleevac and Sprycel are curative.

Careful genotyping may be required, the example is the latest good results of Pfizer study of PF-2341066, a compound originally discovered at SUGEN as a potent c-Met inhibitor (c-Met is anti-metastasis target). It has somewhat dirty profile, and works also on NMP-ALK aberrant kinase that drives a subset of lung tumors. The compound is now doing really well in phase III of non-small cell lung cancer: in patients that have ALK positive cancer type. (But the compound was disappointing in general head-and-neck tumor study.)

Pricing/coverage: cancer drugs are taken for relative short period of time, big pharma needs to re-coop the cost of its ineffective and demoralized research, and with a new unique drug it can hold the cancer patients hostage, charging $1000 for one pill that has 50 mg of a very simple active substance and is taken once a day.

Permalink to Comment

15. Medchemer on June 7, 2010 1:27 PM writes...

The patients need to be told their likely prognosis on the drug. I know that not everyone responds the same way to the same drug but if you asked most 80-year-olds whether they want to prolong their lives for a few months with low quality of life, mostly they will answer - No. The answer may be different for a 40-year-old. Although one can always refuse treatment, doctors generally assume patients want every extra day whatever the physically, emotional, or financial cost. If patients understood what they were signing up for, some may choose palliative care with no treatment.

Permalink to Comment

16. Jeffry on June 7, 2010 1:50 PM writes...

Betsy-- I think you're setting up a straw man. Yes, you're right that its prohibitively difficult to frequently scan everybody for all types of cancers. But we can do a lot better than we are.

Spending money & effort to figure out which lumps are worth worrying about would have huge payback. Better blood tests for just a few cancers would have huge payback. Etc. It doesn't have to be all or nothing.

Permalink to Comment

17. okemist on June 7, 2010 2:26 PM writes...

I am currently working on a resupply of an oncology compound that finished phase II over a year ago. How are you going to find cures if you don't look? What if it extends life for 2 mos in 50% of the population but cures 5%. I do this job for the chance to improve peoples lives, if I look at my success rate, I'd quit. If I keep working for that chance I may succeed. In the end we're all terminal.

Permalink to Comment

18. retread on June 7, 2010 2:26 PM writes...

It's good to see that many of the comments talk about the quality of the extra time being given by the drugs. As a neurologist, usually called in when the tumor had spread to the brain or was compressing the spinal cord, the quality of life in those last few months can be terrible.

Too many times, I saw families who wanted 'everything done' for the patient, as a way of expiating their guilt for the way they had treated them during life.

There is also the conundrum of equating finding cancer early as in some way beneficial. It's certainly logical when you think about it initially. This has been done for several types of tumors, BUT, unfortunately, in several cases the actual mortality figures of the type of tumor did not drop (even though 1 year and 5 year survival figures did increase).

Permalink to Comment

19. Bob Costus on June 7, 2010 2:32 PM writes...

...if we stop paying for it, the companies that make those drugs will disappear...

Is it really that black and white? Is that assuming the companies are charging the minimum value for the drug necessary for them (the companies) to survive? Or are they charging as much as the market will bear, profiting from the human-lives-are-priceless ideal? I think the latter.

Does that profit fund the pipeline for other drugs, like Sirtris (ha), or is it being siphoned off by mismanagement? Likely both. Those who argue any price is necessary to keep the pipeline going have to explain the recent layoffs in R&D at major pharmaceuticals. Apparently the pipeline is not really the most important consideration to management. I'm not aware of any evidence that merging large pharmaceutical companies produces any net benefit on the sum of the pipelines--much evidence to the contrary--but it happens repeatedly anyway.

With regard to price pressures, consider this piece about the recently approved Amgen Prolia from here:

"We view this positively as it appears that Amgen was pricing Prolia with the oncology indications in mind," wrote Citi analyst Yaron Werber in a Wednesday morning research note to clients. "Using this price, the annual cost for the cancer indications (once monthly at double the dose) translates to $19,800 vs. our projection of $12,240." Werber rates Amgen a buy with a $64 price target.

Which has more effect on Amgen's management: the ripple effect of the higher price on the cost of health care, Congressional hearings into the high costs of health care, or a stock "buy" recommendation from a major bank analyst?

Permalink to Comment

20. hn on June 7, 2010 2:32 PM writes...


Gleevec/Sprycel is not curative for leukemia. They convert a terminal illness into a manageable chronic condition. The pills are taken daily indefinitely, at the cost of tens of thousands of dollars per year.

Permalink to Comment

21. JasonP on June 7, 2010 2:37 PM writes...

I'd just like to point out that a bulk of the medical device field developing new instrumentation is geared exactly towards detection of cancers. I guess we're just not there yet.

Permalink to Comment

22. Karen on June 7, 2010 3:04 PM writes...

Another issue is that some of these treatements are moving the bar. So they might only give the patient a few months past the best possible treatment at the time - but a few years down the road, this will be the new standard, and the new treatment will move the bar a few more months down the road, and on and on. If you look at just one treatment in isolation, it may seem like you're only gaining a few months, but if you look at treatment options today compared to 20 years ago, in some cases you're gaining years. Many childhood cancers are treatable now, not because they discovered one magic cure, but because each treatment pushed the bar a little bit further.

Another big issue is that these numbers are averages. Some people may get large gains and others may get zero, and it averages out to an extra 3 or 4 months. We need to figure out who gets the big gains and focus in on those patients. Right now, we have no idea.

Permalink to Comment

23. qetzal on June 7, 2010 3:06 PM writes...

@HelicalZz (#11) -

You raise an interesting point. Can you point to any data? Do you know of any drugs that were approved based on relatively 'meager' benefits in terminal patients, but were later shown to have much more substantial benefits in earlier stage patients?

Even better would be a broader review of multiple oncology drugs from this perspective, though that might be a taller order.

Permalink to Comment

24. billyziege on June 7, 2010 3:54 PM writes...

I echo Karen's comment about raising the bar. Additionally, I'm a little surprised that no one has mentioned the "side" benefit of understanding that may come out of developing a drug in the discussion of cost/benefit analysis. First, publications increase the general body of knowledge that may be applied to other problems or even completely different sorts of areas of research. Second, these endeavors also educate individuals involved who then have experiences they can apply to other systems and problems. While these benefits probably cannot be easily quantified, these benefits should at least be discussed.

On the other hand, one may argue that this understanding could be obtained by appropriating less funds to pure research, but I'd argue that industry provides a certain practical worth to its research endeavors. Also, there are certain political realities of publicly funding projects the size drug companies routinely take on. Of course, I may be biased with my purely academic bend and my lack of knowledge of the quantity and scientific worth of proprietary information that companies may be keeping.

Permalink to Comment

25. Anonymous on June 7, 2010 5:07 PM writes...

"NICE is good in theory but I'd rather money was saved by cutting the huge number of NHS management": why would anyone view these as alternatives?

Permalink to Comment

26. Jose on June 7, 2010 6:50 PM writes...

2.5 months of extra is most certainly worth something to most people- *assuming* they have good quality of life, which is pretty unlikely with stage 4 cancers.

Population based screening is an even more expensive, dangerous option.... you need an insanely good test ($$), both to keep false positives ($$ for followup) and false negatives ($ for treatment later) close to zero. Not very likely.

Permalink to Comment

27. provocateur on June 7, 2010 6:59 PM writes...

dr. Lowe
You got to give your take on this...

PR Newswire - Mountain View CA -

Biotechnology company Vivus Pharmaceuticals announced today that it is beginning clinical trails of a new formulation of its anti-obesity drug Qnexa. The new drug dubbed Qnexa-CM is designed to help patients tolerate adverse effects associated with high dose Qnexa and adhere to the treatment regimen.

CEO Leland Wilson explains "We saw the high dropout rate with Qnexa as a potential problem. So we went back to the drawing board and re-formulated the drug. We've replace the amphetamine Phentermine with Methamphetamine, commonly known as 'crystal meth' - hence the new name Qnexa-CM. We strongly believe the new formulation has several benefits: 1) It's addictive - so patients who are put on the treatment STAY on the treatment. 2)It's effective - patients really lose a lot of weight fast! and 3) It's covered. - After a titration period to start the therapy, patients will need a methadone regimen to end the therapy - methadone treatment to fight drug addiction is covered by most health insurance plans so Qnexa-CM will likely be covered as well."

"The new formulation strengthens our intellectual property position and we plan to pursue worldwide patent protection for our new CM-intermediated metabolic disease treatment technology platform." Wilson continued.

Vivus indicated that the new formulation would be added as a supplement to the Qnexa NDA.

Permalink to Comment

28. Mr AtoZ on June 7, 2010 7:43 PM writes...

Here is another tidbit to add to the morality question once again imposed by a disfunctional government. Estates are exempt from federal tax in the year 2010 but resume in 2011 for the first million. So if you prolong your relative's life even one day to January 1, you get slapped with a Federal death tax. Let him die the day before, you're scott free.

Permalink to Comment

29. Ano on June 7, 2010 8:54 PM writes...

@John, "An interesting aside is that estimates by the National Academies report was between 65 and 70 percent of cancer being preventable or environmental in it's cause."

I've heard this bandied around quite a bit and it strikes me as highly disingenuous. How do they know? Isn't cancer mainly a disease of old-age? Isn't this just another element of a culture which always looks around to blame someone or something, every time a bad thing happens?

Permalink to Comment

30. hibob on June 7, 2010 9:48 PM writes...

@Jeffrey - You should consider the cost of more cancer diagnostics in human terms: increased cancer due to X-Rays, false positives, unnecessary surgeries ...
Which diagnostics in particular do you think the US has been skimping on?

Permalink to Comment

31. Anonymous BMS Researcher on June 7, 2010 9:58 PM writes...

A few points relating to the above comments:

1. As was noted above, companies usually test their newest oncology drugs on incurable cancers first because that's not only the most ethically acceptable population for the first tests, but also the fastest way of getting it on the market. In most other disease areas we start with a small number of healthy volunteers, but the side effects of most oncology drugs are too severe for that.

2. As was also noted above, chemotherapy is frequently done with combinations of drugs; it can take a long time to figure out how best to use a new agent.

3. All pharmas are looking very hard these days for biomarkers that can help predict efficacy. Not easy.

Permalink to Comment

32. Anonymous BMS Researcher on June 7, 2010 10:22 PM writes...

And a fourth point I forgot to mention: the number one problem with many screening tests is the FALSE POSITIVE rate. Suppose you invent an incredibly accurate screening test: if you have a particular type of cancer there is a 99% chance the test will say you have it, and if you don't have that type of cancer there is a 99% chance the test will say you do not have it. But if only one out of 1,000 people given this test actually has that kind of cancer, then this test will give about 10 false positives for every case of cancer detected. This can be calculated with Bayes' Rule for Conditional Probability, or with a little simple arithmetic:

Suppose 100,000 people get the test, out of whom 100 have the specific type of cancer it detects. So out of these 100 you get 99 true positives and one false negative. Out of the remaining 99,000 you get 98,010 true negatives and 990 false positives.

Cancer is not one disease, but hundreds of them. Most of them are quite rare. And of course real screening tests have error rates a lot higher than 1 percent.

Therefore, even if we developed implausibly good screening tests for every known type of cancer, and even if the cost of the test itself were low enough that we could afford to give every test with every checkup, there is no way on Earth we could afford to follow up on the zillions of false positive results we'd get -- not to mention the psychological stress for the patient. My wife went through such a false alarm about 10 years ago, and I can tell you it was a very stressful few weeks before the biopsies came back.

Therefore, screening tests are most practical when (1) the condition is common, (2) the test is cheap, and (3) the condition is likely to be treatable when it is detected. Blood pressure is an example of a screening test that does make a lot of sense: high BP is common, the test is simple to do, the damage from high BP takes many many years to have an effect in most cases, and there are good treatments available. And we still undertreat high BP even though the entire medical community has known about it for decades.

Permalink to Comment

33. processchemist on June 8, 2010 1:44 AM writes...

Having worked with some companies on oncology projects, I totally agree with Anonimous BMS researcher. Sure, there's a lot of "just-another-kinase-inib" around (and some totally improbable stuff). But also other compounds, out of the mainstream, and other projects targeting products with lesser side effects. I don't think that, for many patients, we can say that the current standard of care it's the best thing we can think of. And in my humble opinion biologics will not solve all the problems in the best way.

Permalink to Comment

34. Skeptic on June 8, 2010 3:24 AM writes...

"All pharmas are looking very hard these days for biomarkers that can help predict efficacy"

Ha. Only because