How We Do Harm (27 page)

Chapter 18

Faith-Based Medicine

WHY DO
MY COLLEAGUES IGNORE
science?
Some do it out of ignorance.
Some do it out of greed, some do it out of a weird apathy.
This last group is composed of people who are satisfied with not knowing, with not being informed.
They are not technically ignorant.
They know what they don’t know.
They know there is something to learn.
They just don’t want to learn it.

And who can possibly hold us accountable?
Certainly not the consumers who have come to trust doctors, believing naïvely that our medical care is the best in the world.

Recently, at a congressional hearing, I testified alongside a colleague who threw into his testimony a bunch of statements that anyone who has taken a beginner’s course in epidemiology and biostatistics would have recognized as nonsense.
In one of the most egregious examples, he claimed that an increase in survival in prostate cancer is evidence of effectiveness of screening.
That’s just wrong.
Survival measures time that elapses after diagnosis.
By diagnosing a cancer earlier you by definition increase survival.
Alas, you do not necessarily make the patient live longer.
If you have a cancer with a lot of slow-growing and inert tumors, you can increase the proportion of patients surviving longer than five years.
This is like adding people who do not have cancer to the group.
The more you diagnose, the more you push up survival.
To get something meaningful, you have to measure mortality—actual deaths—from causes related to the disease.

You can explain this to a smart third-grader.
But none of the legislators detected the scientific nonsense.
I called my colleague on this publicly, as a matter of principle, out of frustration.
This didn’t resolve the matter, but was taken as a case of experts-disagree-on-an-obscure-point.

The proper study to show that prostate screening saves lives or averts death from prostate cancer is an expensive study that requires time.
The National Cancer Institute was doing the study at the time of my testimony, but it had not been completed.
Interestingly, a lot of people in the urology community criticized the study.
They would ask, why do it when it is obvious that screening saves lives?
Eventually, the American study would not show that prostate-cancer screening was beneficial with about ten years of follow-up.
A parallel European study with lots of design flaws would suggest a small benefit to screening.
However, that study also showed you had to treat forty-eight men to save one life.
The pro-screeners criticized every aspect of the American study and praised the European study.

They failed to mention the study showed that forty-seven men would receive useless treatment that impacts quality of life and could actually die in order to save one life.
That’s a high cost.

Antiscience in urology isn’t limited to diagnosis.
When treating prostate cancer, urologists are fond of prescribing hormonal treatments.
Often, this treatment is prescribed to men who haven’t been shown to benefit from this treatment.
In some cases, hormones for prostate cancer are just fine.
The therapy—gonadotropin-releasing hormone agonists—is approved by the FDA for palliative care of advanced disease.
Basically, it can be a reasonable final treatment a man gets for this disease.

Some randomized-trial data show that hormones slightly improve survival for clinically advanced localized disease when combined with radiation therapy.
That’s not on the label, but it’s a reasonable use, too.

These, alas, are the only appropriate uses.
No randomized-trial data points to improvement in survival in any other setting.
Yet, when we look at patterns of care, we see that one in three men treated for prostate cancer receives these drugs at some point in their disease.
This adds up to 60,000 to 70,000 new patients per year.
Altogether, at least 250,000 men receive these drugs, paying $800 a month, often for the duration of their lives.

In the absence of scientific data, I can think of only two compelling reasons to prescribe hormones to people who haven’t been shown to belong to the groups that stand to benefit from them.
One of these reasons is supply, the other, demand.

Demand is where we will begin.
It materialized in the 1990s, with the start of screening for prostate cancer with the prostate-specific antigen assay.
Men were being diagnosed by the millionss.

Consider a common scenario.
A man is diagnosed through PSA and his prostate is taken out.
However, after a couple of years, his PSA score begins to rise.
He has no symptoms of progressive disease, just a rising lab value.
For years, urologists have had powerful incentives to put this gentleman on hormones.

It’s logical: prostate cancer is sensitive to hormones; pushing the patient into a hypogonadal state should lower his chances of a recurrence.
The result feels like a win-win situation.
The treatment knocks down the PSA, the patient is less worried, and the urologist is paid.
The problem is, no one can say with certainty whether the patient was helped or harmed.
While he might believe that his cancer was detected early and subsequently cured, in reality he could have been rendered impotent and put in diapers unnecessarily and is about to be finished off with a stroke.
Is it A (a triumph) or B (a disaster)?
No one seems to be interested in testing this intervention in clinical trials.

Through the 1990s, physician-reimbursement policy made administration of hormones profitable to urologists and medical oncologists.
Studies show that reimbursement for hormonal treatments accounted for 40 percent of all Medicare payments to some urology practices in the late 1990s.
Some of these urologists are well-known, and many are revered by the prostate-cancer-treatment advocacy movement.

Total Medicare costs for this class of drugs in 2000 hovered around $1 billion.
At that time, nearly half of all men with prostate cancer were getting one of these drugs.
For years, one pharma company gave free “samples” to doctors, who then injected them in patients and billed Medicare.
The scheme put several urologists behind bars.

These drugs are rough on patients.
Men get hot flashes, headaches, osteoporosis.
Worse, there is diabetes and cardiovascular disease, including heart attacks, strokes, and sudden death.

It would have been nice to have ongoing randomized trials of the intervention, perhaps randomizing men to receive the treatment immediately after their PSA starts going up or after they develop actual symptoms.
Basically, half the patients would receive treatment for indisputable disease, while the other half would be treated for a rising lab value.

Similar trials were conducted with DES, diethylstilbestrol, about four decades ago.
Then, urologists learned that it’s better to start treatment after the patient’s symptoms show up.
Of course, DES was cheap; the new-generation drugs are horrendously expensive.
(If money is important, it’s better to start sooner and sell more drug.)

A trial with proper informed consent and careful monitoring of adverse events would have been great.
Alas, urologists have thwarted all efforts to start such a study.
Instead, by giving these drugs without justification to millions of men who may not benefit from them, urologists have staged a frightening societal experiment.

The impact of this societal experiment can now be seen in registries that track cancer statistics.
These consequences have been documented in three large registries, a finding that led the US Food and Drug Administration to add a warning to the drugs’ labels.
Of course, the agency merely urged caution.
If urologists and their patients don’t want to heed warnings, they don’t have to.

To understand the magnitude of this societal experiment, consider another point of information: deaths from prostate cancer have been dropping every year over the past two decades, and this drop has added up to about 30 percent since 1990.

Is cause-specific mortality dropping because doctors are finding earlier-stage disease and treating it effectively?
That would be the best-case scenario.

The worst-case scenario is at least equally compelling: widespread use of hormonal agents is causing men to die of cardiovascular disease and diabetes
before
they would ordinarily die of prostate cancer.
That’s what I suspect is taking place.

If urologists stop prescribing these drugs as widely as they used to, we will see deaths from prostate cancer start to inch up.
That could actually be
good
news.
Some of the men who would have been killed earlier by strokes and heart attacks caused by hormonal treatments of their asymptomatic disease would now be living long enough to die of their prostate cancer.

Inappropriate use of these drugs can be attributed to the profit motive.
A recent study of prescribing patterns demonstrated that as soon as the profit motive weakened, inappropriate prescribing of these drugs dropped.
Researchers from the University of Michigan focused on the impact of the decision by Medicare to restrict coverage of hormonal treatments.
The changes occurred in 2004 and 2005.
How did these changes affect the treatment strategies?
To answer this question, the researchers analyzed the Medicare and NCI databases to determine changes in practice patterns that coincided with the change in reimbursement policy.

By going through the data, they identified the men for whom hormonal treatment was appropriate and those for whom it wasn’t.
Altogether, researchers went through data on 54,925 men who received a diagnosis of incident prostate cancer between 2003 and 2005.
They determined that a large number of men whose disease was not appropriate for hormonal therapy nonetheless received it during the era of high reimbursement.
However, after Medicare changed its coverage policies, inappropriate usage declined.

The rate of inappropriate use of hormonal treatment dropped from 38.7 percent in 2003 to 30.6 percent in 2004 to 25.7 percent in 2005.
This change in Medicare coverage did no harm.
Appropriate use of hormones remained the same.

There is no cause for rejoicing.
The study suggests that one in four men getting these drugs is getting them inappropriately.
The wages of sin have declined, but they are still pretty good.

*

SOMEWHERE
along the way, we have been conditioned to believe that a newer treatment is always a better treatment.

Consider the case of Prilosec, generic name omeprazole, a blockbuster drug for AstraZeneca.
The pharmaceutical company marketed it as a treatment for ulcers and gastroesophageal reflux disease.
In gastroesophageal reflux, also known as GERD, the backward flow of acid from the stomach causes heartburn and injury of the esophagus, which is called erosive esophagitis.
Prilosec was FDA-approved for these indications in 1989 and, by 1991, was one of the most commonly prescribed drugs in the United States and one of the most profitable drugs in the pharmaceutical industry.

When the patent allowing exclusive production and sale of Prilosec was several years from ending, AstraZeneca executives confronted the inevitable question “What will the next blockbuster be?”
How do they keep the money flowing?
There was a desperate search for a new drug.

After failing to find their next blockbuster drug, AstraZeneca chemists and lawyers came up with the next best thing.
Prilosec was composed of two compounds—left-and right-handed enantiomers of omeprazole.
The synthesis of omeprazole created two compounds—both with the same chemical formula, but different.
Three-dimensionally, the enantiomers are mirror images, just as humans’ left and right hands are mirror images of each other.

Every pill of Prilosec was 50 percent left and 50 percent right.
The right-handed structure had little activity as a drug; most activity was in the left-handed enantiomer.
The liver converts some of the inactive right-handed structure to the active left-handed structure.
Clever chemists at AstraZeneca were able to separate the two isomers and put the active left-handed compound into a new pill, and equally clever patent lawyers at AstraZeneca successfully argued that they should get a new patent for the pill that was only the left-handed enantiomer.

With the new patent, the company conducted a Phase III trial comparing the old omeprazole with the new drug, which was given the generic name esomeprazole.
How would you test such a drug?
After all, you wouldn’t expect it to be better than Prilosec.
The company launched a “noninferiority” study, seeking to demonstrate that the new drug was essentially equivalent to its predecessor.
This got esomeprazole approved for the same uses as omeprazole.

The AstraZeneca marketing group got on the job.
They set out to make the new drug their next billion-dollar blockbuster.
They had the chemists formulate the esomeprazole as a “big purple pill.”
When you set your sights on a blockbuster, names are important, so the company called this Nexium.

The name seems like a jab at us consumers.
The company started advertising the big purple pill in a competition against Prilosec.
Of course, AstraZeneca marketed both drugs that did the same thing.
That Nexium was new and more expensive created, for doctors who don’t follow the literature carefully and to patients who watch TV, the perception that Nexium was better than Prilosec.
Keep in mind, the clinical trials AstraZeneca presented to the FDA showed the two were equivalent.

AZ marketers began successfully destroying the market for the drug they were about to lose the patent on and converting Prilosec users to Nexium.

Sales of Nexium took off.
It would indeed become the next big blockbuster for AstraZeneca.
Eventually, Prilosec would be approved for over-the-counter sales, and AstraZeneca would capitalize on this as well.

People who pay their good money to see a doctor want a prescription.
They do not want to be told to go buy an over-the-counter drug.
The prejudice is that prescription drugs are better than over-the-counter drugs.