Imagine there is no regulatory agency saying which drugs are efficacious, and consequently, which drugs could be marketed. (Assume that safety is still regulated.) Pharmaceutical companies would develop and manufacturer medicines and sell them to patients. If the patient felt better or got better, he/she would keep taking that medicine. If the patient didn't get better, had a tolerability issue, or didn't see the value, he/she would stop taking that medicine.

The problem with this, as we all know, is the placebo effect. The patient's perceived net benefit is equal to the real benefit plus the placebo effect. There could be a real effect like reduced blood pressure or increased CD4 cell counts--or there could be none at all. The placebo effect could convince the patient that there's a real benefit even when there is none.

The FDA's approach to get around this problem is to separate the real effect from the placebo effect via clinical trials with large numbers of patients to determine whether the medicine is really working or not. Sometimes this divide-and-conquer solution isn't possible, or even necessary, when the entire patient experience is holistic and subjective.

Consider pain. In clinical trials, the visual analog scale is used frequently to assess a patient's perceived pain level. This is merely an objective measurement of nothing other than a subjective assessment. A patient's assessment of pain is highly individualized and situationally dependent.

Roche already has a diagnostics division, so they don't need the acquisition to help drive any of their targeted medicines. After all, once you know the "target" for a personalized cancer drug (like Herceptin) and get it on the FDA-approved label you don't need to know anything else about your patient's genome.

So why the Illumina bid? (Besides the fact that the stock is way off its high.) This Bloomberg Businessweek article gives a lot of good background on the bid, and asks a lot of good questions.

Analysts also point out that the market for the expensive gene-sequencing machines - primarily academic scientists with government grants - is a shrinking market right now, so Roche's bid has got to be about the future market for genomic technologies more than the present one.

What is the next market for super-fast, cheap gene sequencing? It's hospitals, doctors offices - heck, maybe even the CVS drug store down the street. That's the future of genome sequencing: fast enough and cheap enough to become a consumer commodity.

(I think that Roche is betting that if you're willing to pay $500 or $600 today for a tablet to play Angry Birds, you'd pay the same - or more - out of pocket to know your or your children's genetic future. For instance, what diseases to watch out for, what drugs or vitamins to take - or avoid - etc.)

The problem I see is that we don't have a health care system, or a regulatory system, that is prepared to interpret the flood of genomic information from Illumina's superfast machines and then turn it into actual clinical knowledge. The FDA has already signaled that it's very leery about consumer genomic services, and without that approval the technology isn't going anywhere. (And even then, it still has to be translated into plain English for physicians and patients.)

Roche, I think, has the complete play here. They're intimately familiar with the regulatory hurdles at the FDA, and know how the agency thinks and what kind of data they will be looking for in terms of regulatory approval for genomic applications. They've got marketing channels into physician and hospital offices, and the science research base to help translate emerging genomic discoveries into clinical information and - better yet - personalized treatments coming out of Roche's labs.

If personalized medicine is going to expand beyond specialized cancer treatments, companies like Roche will lead the way since they have all the tools to translate the genome into mainstream medicine.

The question is, how long will it take (5, 10, 15 years?) for the transformation to become complete, and how much (or how little) regulators will slow the revolution down - in the name of protecting consumers from themselves.

Hopefully, innovative companies will be allowed to lead the way, with the FDA just validating the underlying methodologies.

That's the title of a provocative new book by Dr. David Agus and the topic of an excerpt in the Wall Street Journal this past Saturday, called The Doctor in Your Pocket. In the not too distant future, Agus believes, his children - and everyone else - will be able to:

....monitor and adjust their health in real time with the help of smartphones, wearable gadgets--perhaps like small, invisible stickers--to track the inner workings of their cells, and virtual replicas of their bodies that they will play much like videogames, allowing them to know exactly what they can do to optimize every aspect of their health. What happens when I take drug x at dosage y? How can I change the expression of my genes to stop cancer? Would eating more salmon and dark chocolate boost my metabolism and burn fat? Can red wine really lower my risk of heart attack?

What is equally exciting is that this patient data will be added to a universal database that can be aggregated by powerful search engines like Google and constantly fed into new trials and experiments--speeding up our understanding of which drugs work best for which people. The database might show, for example, that people with a particular genetic profile respond to one type of cancer treatment but not another. As more people anonymously add their health data, this database would become more and more effective as a tool for preventive medicine.

Dr. Agus could've just as easily called the book, The End of Health. Today, we only think we're sick when the flu or the cold virus sends us staggering back to our bed, or the cancer becomes a lump we can touch or see on an X-ray.

In reality, your body is a constantly shifting molecular battleground, with life and death battles being waged every minute by myriad protective genes and the immune system to neutralize invading infections induce pre-cancerous cells to commit suicide, and keep a healthy balance between thousands of other protein-protein interactions.

Typically, tumors start from a single cell a decade before the fatal cancer is unleashed. The first insidious tendrils of Alzheimer's in the brain may launch decades before the dementia becomes detectable. And the diseases that will kill or cripple us are almost uniquely personal to our genetics, diet, and environment, requiring an equally personalized approach to treating the complex diseases that afflict modern humanity. (My colleague Peter Huber has written eloquently on the challenges and opportunities of personalized medicine in a seminal City Journal article, Cherry Garcia and the End of Socialized Medicine.)

So fighting the battle against disease, and the ravages of aging that often cause disease, is happening in your body now, and the drug that the FDA approves to fight the disease after it is half way to killing you is often too late to do much good -and at enormous financial cost.

If we want to really conquer diseases like cancer, as opposed to just slow it down at the margins, we need to mine the information on those myriad interactions in your body in something approaching real time. It'll be tremendously challenging, but as Argus points out all of basic technology - smartphones, supercomputers, whole genome scans, etc., are all available today and getting cheaper all of the time.

What we really need now, is a vanguard of people - hundreds or thousands - to put the technology to the test, and then let the supercomputers loose to crunch the data and uncover the associations that will help drive new prevention efforts to attack these diseases at their molecular roots when they're still just chemicals as opposed to solid tumors or ravaged neurons.

Thankfully, capitalism is very good at scanning things, crunching numbers, and delivering personalized recommendations (a la Amazon) simultaneously to millions of people.

To paraphrase Glenn Reynolds from Instapundit, "(much) faster please."

The $1,000 genome - the capacity of a gene sequencer to read your entire genome at a consumer friendly price - has long been thought to be a tipping point for personalized medicine. And, according to the Wall Street Journal, the $1,000 genome is either here now or will be very shortly.

Once it's cheap enough for millions of Americans to have their genomes sequenced, the demand for personalized drug treatment or prevention regimens should skyrocket. Researchers will also benefit from a diverse array of genomes to scan for links to common and uncommon diseases. Once those links are established (and more are being established ever day) companies will have a self-identified market for targeted therapies. Finally, drug development will also be transformed.

Today, drug development is an empirical, trial and error process that occurs over many years and three distinct phases of FDA-mandated clinical trial testing. Phase I trials look at basic safety testing in healthy volunteers; Phase II trials occur in small groups of patients with the disease in question, hopefully showing both a promising efficacy signal and continued safety. Finally, Phase III trials (pivotal for FDA approval) are the largest trials (with hundreds to thousands of patients) and are conducted in two independent, double blind, placebo controlled trials.

But if you have a targeted therapy with a validated link to a gene variant, you should be able to go from relatively early safety and efficacy testing without the need for large, lengthy, and expensive Phase III trials. This would allow important new medicines to reach patients years faster than with a non-targeted therapy.

This is because whole genome sequencing would allow companies to essentially pre-screen "non-responders" (who don't have the gene) out of their trials. Not only would this save in trial recruitment costs (assuming whole genome scans are incorporated into electronic health records), targeted therapies typically show much higher efficacy than non-targeted therapies, giving the FDA the confidence to approve targeted medicines faster, with knowledge that the treated group will get more of the benefits - and fewer of the risks - associated with testing or prescribing medicines on an all-comers basis.

This approach - allowing medicines to be marketed after Phase II testing in return for enhanced postmarket surveillance - is called conditional approval and its being seriously considered in Congress and at the FDA.

In a nutshell, the $1,000 genome will transform medicine - provided that the rest of the U.S. medical system and the FDA catch up with consumer's ability to access genomic data.

For instance, how accurate will the coming $1,000 genomes be? Even a small error rate could have profound consequences, especially if it led doctors to prescribe powerful drugs based on a whole genome analysis that later turned out to be flawed. And just a few bad outcomes could slow adoption of the technology.

Will physicians be capable of actively interpreting the information and incorporating it into the patient's plan of care? If physicians can't act on whole genome sequencing in a medically meaningful way, the patient has only purchased a $1,000 paperweight.

Can patients and physicians rely on research establishing gene (or gene panel) disease associations? Establishing a link between, say, heart disease or diabetes might encourage a physician to be more proactive in managing a patients' cholesterol or diet - but if the link isn't valid, the added spending or stress won't be worth it.

Finally, will the FDA be willing to shift from its current premarket focus to a postmarket, monitoring approach? In the cancer space, they've taken this approach through the accelerated approval pathway, which allows drugs to be marketed based on preliminary efficacy evidence, followed by additional postmarketing commitments from companies. With the increasing use of whole genome sequencing and electronic health records, the agency should be more comfortable with shifting to a conditional approval approach for more common diseases - but the agency might also require Congressional legislation to prod it to catch up with new technologies.

The first $1,000 genome will be a milestone. But as is often the case with any truly new technology (think IT technology in 1990s) it make take some time for it to be used most effectively. Here's hoping that they work out the kinks very quickly.

The approvals this past August of Pfizer's lung cancer drug Xalkori and the joint Roche-Plexxicon melanoma drug Zelboraf were seen as a hallmark of the coming era of companion diagnostics, in which drugs that treat specific patient sub-populations are approved at the same time as the tests that help doctors determine which patients will benefit from the medicines. But now that the concept has been validated, some drug firms and patient advocates worry that the FDA may be setting too high a bar for important new drugs to be approved. As discussed in a New York Times article last week, the FDA is now beginning to reject drugs that work in some patients but not others when an approved companion diagnostic test isn't available to identify who may benefit and who won't.

Because many drugs are not fully effective for all patients, the practice of medicine can often be a hit or miss affair in which doctors cycle through various drugs in a particular chemical class until their patients respond appropriately. Individual cholesterol-lowering statins, for example, tend on average to work well in just 30 to 70 percent of patients. And individual blood pressure-lowering ACE inhibitors tend to work in just 10 to 30 percent of patients. But more often than not, if the first prescribed drug doesn't do the job -- or does it with unacceptable side effects -- another drug in the class will work quite well. In other cases, though, as we saw with the breast cancer drug Avastin, some patients will respond to the sole drug in a class extremely well, while others aren't helped at all.

The identification of relevant biomarkers that can predict individual patients' likely response to particular drugs, has been a huge medical breakthrough. Zelboraf works in roughly half of melanoma patients, and they can be identified by the presence of a particular genetic mutation. Xalkori works in about 5 percent of patients whose tumors have a particular chromosomal abnormality. And those drugs were approved along with diagnostic tests that identify the relevant biomarkers. The availability of a companion diagnostic test means that those patients who have little chance of benefiting from a particular treatment don't have to waste valuable time experimenting with a drug that could do more harm than good.

But here's where the story gets complicated. Dozens of other helpful biomarkers have been identified, and are now used in diagnosing patients and determining which medications should be prescribed. But, most of the drugs used in conjunction with known biomarkers were approved without needing to wait for approval of a companion diagnostic. According to the Times, "Before August, the only other dual approval was of Genentech's breast cancer drug Herceptin and Dako's test for the related HER2 protein in 1998. There are more than 70 other tests that guide drug use in some way, according to the Personalized Medicine Coalition, but they are rarely required and often developed well after the drug reaches the market."

Unfortunately, very often scientists have not identified a biomarker to tell the lucky few from the unlucky many. As the Times article notes, "Often, scientists simply do not know what to test for to predict a drug's effectiveness, or they don't find out until near the end of the drug's clinical trials." And even when a biomarker can be identified, "coordinating development and approval of a drug and a test -- by two separate companies reviewed by two F.D.A. divisions -- can raise the cost of drug development if not done well." It can also keep important treatment options out of the hands of patients for many years.

In 2010, in what should have been a hugely controversial decision, the FDA rejected the ChemGenex Pharmaceuticals drug Omapro (omacetaxine mepesuccinate), even though a valid biomarker HAD been discovered, because there was no FDA approved or "approvable" companion diagnostic test available to select patients with the Bcr-Abl T315I gene mutation that identifies them as likely to benefit from the drug. According to Richard Pazdur, Director of the FDA's Office of Oncology Drug Products, "The lack of having a uniform in vitro diagnostic test creates uncertainty about patient selection both in this trial and, more importantly, in a post-approval setting. ... If a patient does not harbor the T315I mutation but is falsely identified as having such a mutation by these un-reviewed assay methods, the patient may not receive more effective, less toxic therapy, such as dasatinib or nilotinib. Conversely, patients with a false-negative test result would receive an ineffective therapy."

In effect, the promise of biomarkers and companion diagnostics to help doctors provide the right treatment for the right patient at the right time can be held up if a drug company isn't able to enter the diagnostics business or convince a company already in the diagnostics business to seek approval for a companion medical test.

While I fully appreciate the possibility that a faulty diagnostic test can have severe negative consequences for patients, the FDA's decision in this case is absurd. I agree with Dr. Ellin Berman, a physician at Memorial Sloan-Kettering Cancer Center and the only member of the FDA's Oncology Drugs Advisory Committee to vote in favor of approving Omapro without the companion diagnostic. She told the Oncology Times, "I think the drug should be approved based on what we heard today. The likelihood of having a standardized test for T315I is, I believe, years in coming." She noted that patients with the T315I mutation and who do not respond to the first line treatment have few other options. "Patients are going to be harmed by this [vote]," she said.

The FDA's failure to provide a lifeline for an identifiable sub-population of patients with a deadly disease and who have already failed to respond to the first line treatment is inexcusable. The development of companion diagnostics will be a tremendous boon for patient health, but we cannot let the absence of an approved diagnostic test stand in the way of promising treatment options.

There's an interesting discussion of prostate specific antigen (PSA) tests in the latest UC Berkeley Wellness Letter. The article critiques the U.S. Preventive Services Task Force's recommendation in October that all men skip PSA tests for prostate cancer screening. The Task Force had never actually recommended PSA screening and was prepared to recommend against it three years ago, but was afraid of the backlash.

The article clearly lays out the strengths and weaknesses of the PSA test. In an editorial postscript, the chairman of the editorial board admits that the male members of the board are split about future testing. "Two won't, several are considering stopping, while the rest will continue to be tested, though with some ambivalence."

Most medicine is ambiguous and the current PSA discussion merely underscores that fact. Even the informed editorial board of Wellness Letter is split on what to do. The final recommendation in the article is: "The decision is a personal one, and men should discuss the pros and cons of PSA testing with their doctors starting at about age 50, earlier if they are high risk."

Let's be honest. This is the recommendation for more than just PSA testing; it is the recommendation for all medical cases. All cases are personal.

Imagine the editorial board's uproar if the chairman decreed that the other members follow his decision, effectively forcing his preferences on them and eliminating the role of individualized treatment and experimentation.

Now imagine that he dictated his preferences to 313 million people and you get a good picture of the FDA. The FDA will never understand the specific circumstances and preferences of 313 million people. The FDA's autocratic approach denies Americans the opportunity to make personalized treatment decisions.

Prometheus gave man fire, thankfully he didn't charge every time man lit a match. Prometheus Labs in contrast wants to charge patients for a rule that says when to increase or decrease a drug in response to a blood test. Quoting Tim Lee:

The patent does not cover the drug itself--that patent expired years ago--nor does it cover any specific machine or procedure for measuring the metabolite level. Rather, it covers the idea that particular levels of the chemical "indicate a need" to raise or lower the drug dosage.

Even this is not quite right for suppose a physician notes that the patient's metabolites are within the range where a change in dosage is not necessary; although the physician takes no action she still has used the patent and thus must pay Prometheus Lab a fee or infringe.

We already have significant incentives for producing pharmaceuticals (and thus the instructions required to best use those pharmaceuticals), we support medical research through universities and non-profit hospitals, and there is plenty of opportunity to profit from the manufacture of tests. Will we really get enough additional innovation to justify the monopoly prices and deadweight losses when we enforce patents on medical rules? Remember, we have to pay the higher prices on all the rules not just the ones brought into being by the patent.

And if medical patents why not economic patents? Will Scott Sumner now patent a rule for adjusting the money supply in response to metabolites the futures market?

Patents like this are a logical consequence of the extension of patentable matter to software and business methods but extending patents to software and business methods has created huge legal costs without any increase in innovation.

Most importantly, patents can reduce innovation and are especially likely to do so in fields where innovations build on innovations. In fields of cumulative innovation, previous patents owners become veto players who can threaten to holdup the new innovation unless they are granted a share of the proceeds. In theory, bargaining can result in an efficient outcome. In practice, it means lawsuits, delay, waste and reduced innovation.

Since a smartphone may rely on many thousands of previous patents, the smartphone industry has heretofore been considered a classic case of how too many veto players can impede innovation. But now consider human metabolism, one of the most complicated systems known to man (just a tiny fraction of that system is shown at right), and note that if Prometheus is successful in this lawsuit that any correlation in that system can be patented. This is a recipe for disaster.

Addendum: Scotus Blog has a roundup of links. See Launching the Innovation Renaissance (Amazon link, B&N for Nook, also iTunes) for more on patents and their problems. Hat tip also to E.D. Kain who writes:

The world, it appears, is determined to turn me into a full-fledged libertarian. What with SOPA, PIPA, the NDAA, software patent trolling, police violence, and now patents on how doctors provide treatment to their patients, it's becoming more and more clear how pernicious the law can be when it's designed for powerful special interests, national security hawks, and big corporations.

Cross Posted from Marginal Revolution.

The United Kingdom's National Institute for Clinical Excellence (NICE) reviews applications to market new drugs in the U.K. based on their cost-effectiveness, and routinely recommends that Britain's National Health Service not pay for (i.e., refuse to cover) medicines that it thinks are too expensive - like Avastin for metastatic colorectal cancer. So NICE Chairman Michael Rawlins (who's led the institute since its foundation in 1999) is certainly no friend of the drug industry.

This made it all the more surprising this week when Rawlins told D.C. based BioCentury TV that the way that drugs are developed today creates "pervese incentives for companies to charge high prices."

"The first perverse incentive is the way that over the last 15 years the regulatory requirements for both the FDA and the EMA have increased hugely," he said.

"It is a huge increase with not much gain, not much benefit from these increased numbers. And of course, it puts up the cost of drug development hugely," he said. By Rawlins' reckoning, clinical trials drive well over half of the cost of new drugs."

This is a point that industry, patients' groups, and even the FDA is beginning to recognize: that the enormous costs and delays associated with drug development has not just financial consequences for the health care system, but needlessly delays patient access to important new therapies.

And much of the information captured in lengthy and expensive pivotal Phase III clinical trials required for FDA approval could just as easily be captured in the post-market environment through electronic health records or patient registries that could confirm efficacy, monitor drugs for rare safety problems, and help make off-label use of drugs much more scientific.

One approach for speeding drug development while still minimizing safety risks is called progressive approval, an idea recently proposed by the Biotechnology Industry Organization, and in a bill from U.S. Senator Kay Hagan (D, NC).

I'll be writing much more about this in future posts, but the basic idea is fundamentally sound, and has been advocated by various groups for years. Indeed, the FDA already does something like this through it's accelarated approval process (for cancer, AIDs, and orphan drugs), and the BIO and Hagan progressive approvals pathway would merely build on this experience.

The rebuttle, of course, is that many drugs that look good in early stage testing fail in Phase III trials, so progressive approvals would merely expose patients to more "placebos". However, it is also true that many good drugs are never developed because of the enormous costs and time required to bring them to market. And many patients die or suffer waiting for effective drugs to be approved.

Even if some drugs wind up being withdrawn under a progressive approvals process - like Avastin for metastatic breast cancer - many more patients would benefit from early access to more effective therapies. And, in a plus for health care payers, as drug development costs decline drug prices should too. Wouldn't that be really NICE?

Today Lipitor, the world's best selling drug and Pfizer's single most profitable product, loses patent protection and goes generic. If Lipitor (generic name atorvastatin) holds true to form for other generic transitions, Pfizer can expect to lose much of its market share and the price of atorvastatin will eventually plummet by as much as 80%. But while everyone is celebrating the "birth" of yet another cheap generic from an expensive blockbuster, it would be good for policymakers and the press to pause and reflect on where generic drugs come from.

The stork doesn't bring them to your local pharmacy. They are often the products of years or even decades of enormously expensive and painstaking research by innovative (and yes, profit-driven) companies.

What is the proper response to Lipitor going generic? Thank God for blockbuster drugs.

Because without those drugs we wouldn't have the tremendous health gains from statins and other widely used drugs to treat stroke and hypertension - or the cheap generics that the health care system gains once those drugs lose patent protection.

Cardiovascular disease has been the leading killer in the U.S. every year since 1900, except for 1918 (pandemic flu?). Mortality peaked in 1950, but declined by 60 percent between 1950-1999. Undoubtedly, much of that decline was driven by changes in population health (i.e, declining smoking rates, etc).

However, as the chart above shows, declines in cardiovascular disease - with the exception of a slight uptick for hypertension - continue in the most recent data. For instance, U.S. age-adjusted mortality rates (per 100,000 population) for coronary heart disease fell by 35% from 1999-2007 (from 194.6 to 126); for stroke by 31% (from 61.6 to 42.2); and by 43% for myocardial infarction (from 73.2 to 41.4). Those gains are not all attributable to statins - stents, better drugs for hypertension, better surgeries and diagnostics all play a role - but medical innovation is a central part of the story.

Those gains are even more remarkable considering recent and rapidly rising obesity rates in the U.S., since obesity is a known risk factor for cardiovascular disease (particularly through complications like diabetes). This massive decline in mortality rates from cardiovascular disease is one of the most astonishing phenomenons of modern American medicine - and yet it goes largely unnoticed.

Fifty-something year-old men dying of heart attacks used to be unremarkable. Now, it is considered a terrible tragedy.

Statins were not developed by the U.S. government, or the NIH. As Zycher, Dimasi, and Milne wrote in a 2008 Manhattan Institute report:

Beginning with the ongoing Framingham Heart Study, which has been conducted by the National Heart, Lung, and Blood Institute of the NIH since 1948, the causal relationship between elevated cholesterol levels and cardiovascular disease has become widely recognized.

In 1976, Akira Endo and other researchers at the Sankyo Company and at Beecham Research Laboratories independently isolated mevastatin from fungi, after having screened more than 8,000 microbial extracts. Further research by Endo and colleagues showed that mevastatin reduced cholesterol levels in the liver; subsequently, Endo and researchers at Merck separately isolated lovastatin from a different fungus.

Lovastatin was shown to be more potent than mevastatin and was the first HMG-CoA reductase inhibitor approved by the FDA, in 1987, for the reduction of plasma cholesterol. Further research by private-sector laboratories has yielded additional statin drugs more potent and/or with fewer side effects than lovastatin: pravastatin, simvastatin, atorvastatin, and others, the newer of which have been synthesized in laboratories rather than isolated from natural materials. Private-sector research, in short, developed compounds exploiting new knowledge of a specific disease process, and developed improvements in terms of potency and side effects.

The development of statins also depended on the passionate belief of industry scientists that they had discovered life-saving medicines, and who championed their development even when bad data suggested that they probably should give up. As Harvard Medical School Professor Thomas Stossel wrote in the Boston Globe in 2008:

When Endo published his findings in scientific journals, other companies started searching for statins. One of these companies was Merck, whose researchers demonstrated by the early 1980s that a statin could lower LDL cholesterol in the blood of healthy volunteers without side effects.

The controversy surrounding Merck's painkiller Vioxx so empowered anti-industry critics that imagining any company (or the FDA) taking similar risks today is difficult, given the toxicity data that came close to killing the industry's work on statins. Critics vilified the same Edward Scolnick and Merck for developing Vioxx after it unexpectedly promoted the very same cardiovascular problems prevented by statins. One has to wonder if medical product companies are shelving potentially life-saving products because industry critics are pushing policymakers and regulators to near zero tolerance for rare adverse events.

Stossel's concern is sadly warranted. Policymakers and the media need to be continually reminded that there is still much about human biology that we do not understand, or understand incompletely, but that ignorance should not be taken to be an indictment of industry or cause to embrace a precautionary principle that penalizes innovation through risk-averse FDA regulations or reimbursement strategies that limit physicians to prescribing cheap generics.

Several years ago, Pfizer lost over $800 million dollars when their attempt to improve on statins by raising "good" HDL cholesterol through a new drug called torcetrapib imploded in a late stage clinical trial. Companies continue to take those enormous risks because of the tremendous good they can achieve when they are successful, and because the financial rewards of success still (at least for the time being) compensate them for their many failures.

Patent expiration - for Lipitor or any other drug - is part of a virtuous cycle of innovation than can take decades. Losing sure cash-cows like Lipitor pushes companies to invest billions of dollars in the hope that they will uncover even better drugs for treating heart disease and stroke, or cure Alzheimer's. Some of these drugs may be blockbusters that treat large populations, and others will be "niche busters" that treat patients who don't respond (for genetic or other reasons) to those blockbusters.

But we must continue to defend and explain the virtuous cycle of innovation if we want companies to find the next public health breakthrough, like statins. Even though, in time, we'll take those miracle drugs for granted too - and cheer loudly when they become cheap generics.

In addition to the great MPT posts by Charley Hooper and Jim Pinkerton and myself, Saturday's Wall Street Journal weighed in decisively against the FDA's decision:

The risks of Avastin are real, but they're also well-understood and manageable, especially in end-stage oncology where there are no good options. The FDA's real goal was to send a warning to the rest of the drug industry about who is in charge of drug development. The FDA withdrew Avastin's breast cancer approval last year--leading to Genentech's unprecedented appeal and a two-day trial in June. ...

All of this suggests that Avastin should remain on the market as one treatment alternative as knowledge about the drug grows--which is all that Genentech requested in its appeal. Looking at the same studies, the European Medicines Agency (the FDA's continental equivalent) continues to approve Avastin for breast cancer. The National Comprehensive Cancer Network, a highly respected consortium of U.S. oncology programs, has four times reaffirmed its recommendation that Avastin is "an appropriate therapeutic option."

Read the whole thing.

On Friday, the FDA announced its decision to revoke Avastin's labeling indication for metastatic breast cancer, granted through the agency's accelerated approval (AA) program. The FDA and the Obama Administration are bound to take some heat for this one (as a prelude to Obamacare rationing) but this is one case where I think the FDA did the right thing by following its own rules for AA.

I wouldn't hesitate to criticize the decision if I thought otherwise. A few quick points:

First, the FDA's decision is a lagging indicator. After the most recent studies on Avastin's use in metastatic breast cancer came out, use of the drug plummeted as oncologists interpreted the data and decided that it wasn't working as well for patients as early data had suggested it would. The market, in other words, is working to incorporate information faster than the FDA. That's a good argument for shifting the agency to being more of an information conveyer than a pre-market gatekeeper.

Second, Avastin isn't going to be withdrawn from the market. Doctors can still use it off-label, as they always could, to treat metastatic breast cancer - and they can make that decision on a case by case basis. Medicare is still covering Avastin treatment for breast cancer, and the National Comprehensive Cancer Network still lists Avastin for use in metastatic breast cancer. (The NCCN affirmed this decision in July.) Private insurers like UnitedHealth, that follow the NCCN, will likely continue to pay for Avastin. Some insurers may stop covering Avastin, but this is a market decision, not the FDA's decision. Don't shoot the messenger.

Third, if you think (as I do) that the FDA should be expanding the accelerated approval pathway and allow more drugs to get to market based on promising early studies. rather than waiting for large Phase III clinical trials that can take years to complete, you can argue that this outcome actually strengthens AA. Critics have charged that AA is sop to industry, and that companies never do the follow up studies to support AA. Avastin proves them wrong.

Finally, this is far from the end of the story. Experts agree that some patients with metastatic breast cancer do very well on Avastin, but no one knows how to identify those patients ahead of time. Roche has already announced that's conducting a Phase III trial looking at a potential biomarker for identifying the patients who are likely to get the most benefit from Avastin.

That's exactly where cancer treatment is going: biomarker driven treatments that are given to the patients who are likely to get the most benefit and the fewest side effects (like Herceptin, a targeted breast cancer drug).

In the future, broad labels like "metastatic breast cancer" will mean less and less as drugs are built from the ground up to attack the molecular mechanisms of cancer. Avastin's use will undoubtedly be further refined and improved through the same process. The FDA's decision today won't stop that, and it might even help accelerate it.

For more background, here's a link to the New York Times article on Avastin, and to Commissioner Hamburg's 70 page memo explaining the agency's decision. You can find the report here.

Some early reactions from the Web:

Derek Lowe thinks that this decision is right on the science, and has to happen if cancer treatments are going to improve.

Avastin doesn't work as well as we thought it did for this indication. If you're going to believe in medical progress at all, you have to believe in what multiple well-controlled clinical trials are telling you - trials carried out, keep in mind, by the drug company that has every interest in having them come out favorably. But they didn't. On medical grounds, on scientific grounds, this was the right decision.

Ed Silverman interviews Harvard Professor Daniel Carpenter at Pharmalot:

"This decision sets an important precedent: accumulated science has trumped politicized argument," says Daniel Carpenter, the Allie S. Freed professor of government at Harvard University and author of 'Reputation and Power: Organizational Image and Pharmaceutical Regulation at the FDA.' "The agency faced enormous political pressure from the company and from organized interest groups, and yet rendered a strong decision based upon a searching reading of the available evidence. ...

"In an ironic way, this decision actually allows the FDA to expand the accelerated approval program, as there is now a strong, visible and clear precedent for withdrawing an approval when postmarket experiments for an approved drug produce disconfirming efficacy evidence and perhaps problematic new safety signals," Carpenter says. "This is not an anti-innovation decision and could, in fact, assist pharmaceutical innovation in the years ahead...I predict its longer-term impact will be to help rebuild the agency's reputation for science-based decisionmaking and for independence from emotionally charged lobbying."

I've been meaning to write about a very interesting Forbes article by Matthew Herper on Bill Gates' views of the pharmaceutical industry. Basically, the Gates Foundation has made a huge investment in vaccine development for major diseases in the developing world, like malaria and HIV, over the last decade. This engagement has given Gates an insiders' perspective on the industry, at least compared to other software or tech industry giants who've weighed in on the R&D challenges facing their pharma cousins.

Herper writes that:

After a conversation with Gates, one wonders if the drug industry has hurt itself by focusing too much on $100,000 cancer drugs and blockbuster pills that can be sold to millions, and not enough on products that are far more cost-effective. And then, of course, there's the fact that the entire $600 billion pharmaceutical industry has spent the past decade in a research drought, getting comparatively few new medicines to market. ...

When it comes to cancer drugs like Avastin and Erbitux, some of the biggest sales successes of the past decade, Gates seems unimpressed. "There's always this divergence between what's financially attractive and what has dramatic profit and the number of life years that you really save." Take for instance, Novartis' Gleevec, the crown jewel of targeted cancer drugs that can put chronic myelogenous leukemia and gastrointestinal stromal tumors into remission. "Do the math on that versus, says, preventing Parkinson's or preventing Alzheimer's. It's in a different universe."

It's hard to criticize a billionaire, but I think Gates misses a very big difference between pharma and the software industry: the FDA.

Microsoft doesn't have to go through any pre-market screening (as drug companies must do with the FDA) before it issues its next iteration of Windows, which is just a tweak on the last version. Generating safety and efficacy data for FDA approval for a new medicine can take upwards of $1 billion and the better part of a decade.

As for Gates' criticism of drug companies' development strategies...well, there's the FDA to think about again. The extraordinary risk and expense associated with drug development dictates what types of products companies can pursue, and still stay in business. Cancer drugs and other orphan indications (like Gleevec) benefit from a number of different accelerated regulatory pathways - Fast Track, Accelerated Approval, Priority Review, etc. that make it easier to get through the FDA. In short, the market has, not surprisingly, moved towards the regulatory path of least resistance.

Gates is certainly right, however, that the industry is competing against itself in the form of many "blockbuster" drugs from the 1990s that are swiftly losing patent protection and becoming cheap generics (like Lipitor will shortly). Public and private insurers are much more likely to grant premium reimbursements for drugs that treat serious and life-threatening diseases even for smaller populations when there aren't other good therapies available.

Still, there is tremendous unmet medical need for many widespread, chronic illnesses, like depression and Alzheimer's. But even in these instances, you've got lots of regulatory hurdles to worry about - and the costs of clinical trials for things like Alzheimer's can be astronomical.

If the costs and time required to develop new medicines could be significantly streamlined, using biomarkers or other tools, we might see an explosion of private investment not only for costly chronic illnesses, but also for the "neglected" diseases that the Gates Foundation focuses on.

This will entail something analogous to an Apollo program for FDA reform - setting ambitious goals and holding the agency (and the industry) accountable for making it happen. Kennedy didn't know how we'd get to the moon in 1960, but by 1969 Neil Armstrong was there. Today, we don't know exactly how we'd make a quantum leap in drug development - develop and approve a groundbreaking AD drug in 5 years - but we know that we need to do it.

It'll take a tremendous exercise of political vision and will to lead this kind of effort, and get the stakeholders and regulators to rally around it. If Gates wanted to leverage his own bet on vaccines, this is exactly the kind of project he should embrace.

Not surprisingly, there's been a lot of commentary about the FDA's recent report touting a near-record 35 new drug approvals in 2011. As Paul Howard recognized, the FDA gladly taking credit for the increased rate of drug approvals in 2011 seems more than a bit disingenuous, especially after blaming drugmakers for the lower number of drugs approved over the last few years (21 in 2010, 24 in 2009, 23 in 2008, and 19 in 2007).

For years, the agency has claimed that the scarcity of new drugs approved was the result of lower quality applications, but now that approvals are up, they are singing their own praises. The FDA commissioner's statement about this apparent contradiction, "I think the point we're trying to make is that when high quality science, good applications come before us, we are able to act swiftly and surely," doesn't resolve this discrepancy. Nor does it address internal process delays that need to be addressed by the agency.

Even so, there are lessons to be learned from the types of drugs approved, rather than the total number:

Of the 35 approvals, two are theranostics -- personalized drugs approved for use in conjunction with a specific diagnostic test; seven are cancer drugs; 10 are for orphan diseases; and a total of 16 were approved under priority review. 2010, in contrast, only saw 21 approvals, and the outright rejection of two potential blockbuster weight-loss drugs. The argument I have been making in my last several posts now appears to be a reality. Increasing opportunities exist for drugmakers to pursue alternative research venues, particularly in the areas of theranostics and orphan drugs.

While historically on the margins of R&D spend, orphan drugs represent the wave of the future in drug development. As I discussed in my recent post, the blockbuster model is wearing thin. The pressure for lower cost drugs with better outcomes flies in the face of the "one-size-fits-all" approach to drug-making. Clearly, pharmaceutical companies must find a different way to be profitable.

Staying competitive in the 21st century means cost-cutting and targeting segmented markets. Cost-cutting has been significant in the industry recently. Pharma companies slashed commercial budgets and pulled back thousands of reps from the street, but they still have to make significant investments in R&D. The question is what and how. Here's where orphan drug development fits in.

This article, on the importance of renewing the Prescription Drug User Fee Act, ran yesterday in the Washington Examiner:

Is America still a place where big ideas can move policymakers to slash red tape in drug development to save lives, drive innovation, and spur economic growth?

It certainly appears so. The Obama administration, industry, the Food and Drug Administration and patients' groups have all come together in recent months to highlight the need for regulatory reforms to help patients gain faster access to safer and more innovative medicines.

Streamlining innovation begins with improving the FDA's framework for evaluating the potential risks and benefits of new medicines -- and making it as transparent, predictable and science-based as possible.

For starters, policymakers should reauthorize the Prescription Drug User Fee Act, bipartisan legislation first passed in 1992.

Renewed by Congress every five years, PDUFA gives the FDA authority to collect user fees from companies when they submit new drug applications. The agency uses those fees to hire and train the staff it needs to review applications in a timely and predictable manner. (Note, however, that those fees have no influence on whether or not the FDA approves a product.)

PDUFA also sets timelines for the review of new products that can help evaluate the FDA's performance.

America's ecosystem for biomedical innovation is in dire need of an overhaul. Today, it can take over a decade and $1.3 billion to bring a single new FDA-approved medicine to patients.

Industry and the National Institutes of Health also spend nearly $100 billion annually on basic and applied medical research, but only about 20 new drugs manage to reach the market each year.

Venture capital for biotech is also being cut back, with investors citing an unpredictable and expensive regulatory process. American patients are also waiting longer for access to some new medicines and medical devices that have already been approved in Europe.

The FDA isn't the only stakeholder responsible for innovation, but it is effectively the "rate-setting" institution, since no medicine or technology can make it to market without first gaining FDA approval.

Yesterday the FDA released a report analyzing the "crop" of drug approvals for 2011 - and it's an impressive list. First, the FDA notes that 35 new drugs were approved in 2011, far above the total of 21 in 2010, and the highest total in the last decade except for 2009.

It is certainly an all star list of new medicines:

Seven of the new medicines provide major advances in cancer treatment.

Two new therapies, for lung cancer and melanoma, are breakthrough products for personalized medicine: each was approved with a diagnostic test that helps identify patients for whom the drug is most likely to bring benefits.

The agency also notes that 24 (70%) of these medicines were approved first in the U.S., and that half were approved under "priority review", a designation that requires the agency to review the sponsor's application in 6 months or less. All but one of the 35 applications were reviewed before their target date under PDUFA, and the majority were approved "on the first cycle".

The FDA says that this bumper crop reflects

...many improvements in FDA's drug approval process in the last several years. The agency has made great strides forward to speed the development and availability of drugs for serious or life-threatening diseases; it has launched the Critical Path Initiative to help streamline drug testing and review; and it has sharpened its focus on methods of efficiently identifying and resolving drug safety issues.

These results should probably be taken with a grain of salt: they may represent products that have been in development for years (or decades) and thus could easily represent an outlier in terms of new drug approvals. We could be back to fewer new drugs next year, and then 2011 looks like a blip on the radar.

Still, there are clearly some good stories here. Pfizer's Xalkori, for late state lung cancer was approved with a companion diagnostic to target the patients who are most likely to benefit, received an FDA "Fast Track" designation (where the FDA meets frequently with the sponsor), accelerated approval designation (which allowed approval of the drug based on a surrogate endpoint, in this case shrinking tumors), and priority review.

Overall, Xalkori was approved just 5 years after the first human clinical trials - nearly three years shorter than median clinical trial times for oncology drugs. Other important advances this year included the first drug to improve overall survival for late-stage melanoma patients.

Undoubtedly, the FDA is looking for some good press after facing a flurry of critcism from medical device makers and the venture capital community. So kudos to the FDA for a good year in 2011. (At least at first glance).

But the better question is, what are they going to do ensure even better trends for patients and personalized medicines next year, and for the next five or ten years after that? More on that in future posts.

Xconomy reported last week that consumer genomics company 23andMe may have discovered a genetic variation that confers protection against Parkinson's disease:

Using genetic data drawn from thousands of 23andMe customers, the company says it has identified a gene that appears to protect against a genetic mutation associated with Parkinson's disease. Specifically, 23andMe says the gene serum/glucocorticoid regulated kinase 1 (SGK1) appears to be protective against a mutation known as leucine-rich repeat kinase 2 (LRRK2).

A specific mutation on the LRRK2 gene, known as G2019S, is recognized as a risk factor for developing Parkinson's. About half the people with the mutation develop the disease. Yet 23andMe says it has genetic data from a large number of people who carry the mutation, but who surprisingly don't have Parkinson's. In scrutinizing this group, 23andMe says it made the first-time discovery of the potentially protective nature of SGK1.

Expect the pace of these types of discoveries to accelerate as the cost of genome scans plummets towards $1,000 (or even $100) and millions more consumers opt to have their genomes sequenced. The potential is enormous not only to discover new therapeutic targets for drugs, or develop new tests for genetic diseases, but to create a wave of consumer demand for early prevention and monitoring to nip diseases like cancer or Alzheimer's in the bud years or even decades before they might become clinically manifest.

That is, if regulators don't strangle the new technology in red tape first.

The U.S. Preventive Services Task Force unleashed a fire storm of criticism recently when it recommended that "healthy men" should stop receiving a PSA test to screen for prostate cancer.

The problem with recommendations like this is that they'll be obsolete by the time the government manages to implement them (which, given the backlash, is extremely uncertain anyway).

It is true that the PSA test has a high number of "false positives" that can drive men who don't have cancer, or don't have life-threatening cancers, to get biopsies or treatments that may have serious side effects but not actually save lives. But this is like saying that the first generation of cell phones was ridiculously expensive, clunky, and didn't work very well - and so we shouldn't invest in better cellphones since the early ones were so crappy.

Since we don't adhere to that philosophy in technology markets, relatively cheap, fast, and powerful smartphones now rule the world.

Medical technology is benefitting from the same underlying forces that drove the telecommunications and computing industries into producing incredibly rapid, consumer-friendly innovations.

For instance, take a new biomarker test from the San Diego company Gen-Probe, which is under review by the FDA. Unlike PSA, which is produced by inflamed prostate tissue (which may or may not be cancerous), Gen-Probe screens urine for a prostate cancer antigen (PCA3) which is "overproduced in more than 90 percent of patients with prostate cancers."

Depending on how the FDA approves the test, it could be used as a second line screen for patients with unusual PSA readings, allowing doctors and patients to make better decisions about when to seek a biopsy or leave well enough alone.

That's what we really want, right? More confidence that we're seeking treatment when we need it, and it might save lives, and less treatment when we don't. And the PCA3 test isn't the only promising technology in the pipeline.

(BTW, Gen-Probe's PCA3 test was approved in Europe in 2006.)

I think what Paul describes here is certainly a trend, but it has its limitations.

If each drug takes $1 billion to reach the market and 10 million people use it over its patent protected lifetime, then each patient contributes, on average, $100 to the development of that drug. If we keep shrinking the denominator, then the economics become more difficult. Taken to the extreme of personalized medicine, with one specific drug for each person, we cannot expect that one person to cover the $1 billion development cost. Even if the development cost drops to $1 million per new drug, the economics won't work.

I think the average development cost would need to drop to $10,000 per drug to be reasonable. To reach this price, we would need to exclude the FDA completely--allow drugs to be marketed without prior FDA approval--or allow the FDA to approve the process of drug development instead of each specific drug. With this arrangement, the FDA would evaluate and approve the process of developing personalized medicines, but would then stand aside and let the drug companies deal directly with patients, physicians, and managed care organizations.

Just like drug companies which built their business models on blockbuster medicines, the FDA will need to rethink its approach in the era of micro-market and truly personalized drugs.