How We Do Harm (12 page)

Procrit and Aranesp bucked that trend.
As supportive-care drugs, they could be used in patients undergoing chemotherapy for any solid tumor.
You could give it to every patient, unless the patient had blood cancer.

Doctors don’t have to prescribe drugs based strictly for uses listed on the FDA label.
They prescribed these drugs “off label,” too, giving them to patients with blood cancers and patients whose anemia was attributed to cancer itself, as opposed to chemotherapy.

Doctors could also prescribe these drugs for anemia associated with kidney disease, AIDS, and surgery.
The market was gigantic.

The treatment of anemia was anyone’s dream franchise, the sort of thing people don’t give up voluntarily.

*

YOU
can entice or bully doctors into a consensus.
You can invent a medical condition called cancer fatigue, stage skits, and play them during the Super Bowl to convince patients that a simple prescription can make their tiredness go away.
You can do all this, but sooner or later, nature will look you in the eye and, after making sure that you are paying attention, give you the middle finger.

Soon after hemoglobin-building drugs were approved, a German radiation therapist named Michael Henke decided to test one of the fundamental tenets of his subspecialty: that patients with higher hemoglobin levels have better responses to radiation therapy.
Though this assumption was untested, it had made its way into textbooks.

Henke believed in the connection between hemoglobin and response to radiation.
His goal in trying to officially prove it was to demonstrate that increasing a patient’s hemoglobin with a Procrit-like drug would bring about better outcomes.
He was such a true believer that the Swiss drug company Roche gave him access to its own hemoglobin-boosting drug and research funding for a 351-patient study, which he started in 1997.
Roche was happy to help out.
A positive study would likely have led to regulatory approvals and expansion of what was already a lucrative franchise.

The study’s results didn’t come out the way Henke expected.

He was shocked.

“I was a strong believer, but sometimes you see different things,” he told a reporter in 2003, after his results were published.
The study showed that patients who received the hemoglobin-building drug didn’t live as long as those on placebo.
Also, the disease progressed more rapidly in patients receiving the drug.

Henke concluded that he had encountered a biological phenomenon: the drug seemed to be encouraging tumors to grow.

When a reporter called an Amgen source to discuss Henke’s findings, the source said that Henke was a gadfly who didn’t understand clinical research, a well-meaning boob.

I didn’t think Henke was a boob.
His findings reinforced my concerns that anemia-building drugs were overused.
Several of my academic colleagues agreed with me.
After publication of Henke’s results, doctors at Emory and at the University of Texas MD Anderson Cancer Center in Houston stopped prescribing these agents to patients with treatable head-and-neck cancer.
However, doctors in private practice or at private hospitals and cancer centers continued to prescribe these drugs.
Sales figures indicate that most docs and most hospitals didn’t dwell on Henke’s important findings and, for the most part, either didn’t pay attention or didn’t care.

In August 2003, researchers had to stop another study, the Breast Cancer Erythropoietin Survival Trial, abbreviated as BEST.
In that trial, more women died on Proctit than on the control arm.
This result, obtained from that massive 939-patient study, was more relevant to Lilla’s case.
In this study, patients received Procrit when their hemoglobin level dropped below 13 g/dl.
In both the Henke trial and BEST, the survival curve showed an increased risk of death from cancer, which suggested something you don’t want to see in patients you are treating for cancer: tumor growth.

This finding was so shocking that the principal investigator of the BEST study, incredibly, cautioned doctors against drawing far-reaching conclusions.

“It is extremely unfortunate that the problems in design, conduct, and post-trial analysis have complicated the interpretation of this study,” the principal investigator, Brian Leyland-Jones, then of McGill University in Montreal, wrote in
The Lancet Oncology
in August 2003.
“Given the number of design issues uncovered in the post hoc analyses, the results cannot be considered conclusive.”

Usually, clinical researchers live and die by their data.
Here, the data were so much at odds with medical practice and prevailing beliefs that Leyland-Jones was urging his colleagues to disregard the evidence and go by his opinion instead.
Never before had I seen such guidance in medical literature.

After BEST, people I respect in breast cancer became cautious about Procrit and Aranesp.
On the pharmaceutical end of things, far from disregarding Leyland-Jones’s findings, J&J decided to stop all studies where the treatment goal was to push hemoglobin into the so-called normal level.
Amgen, on the other hand, continued such studies.
The company’s decision to continue this research was consistent with its overall aggressive marketing stance.
Pushing the cancer patients’ hemoglobin to new heights would have been worth billions in new sales.

The FDA, too, was taking notice, largely because the agency’s head of oncology, Richard Pazdur, is a natural skeptic who concluded that the studies pointed to overuse of these drugs, and their potential dangers.
Pazdur is a fine scientist and physician, but you don’t have to be either to recognize the signs of system failures in medicine.

In the case of hemoglobin-boosting drugs, none of us really knew how to establish uniform hemoglobin targets that would hold for all patients.
We didn’t know what was normal, what the target should be.
We had no idea what dose of these drugs to use and when it made sense to stop pumping it into patients.
If doctors had read the label, they would have learned that the FDA approved these drugs for reducing the risk of blood transfusion in patients with solid tumors treated with chemotherapy.
That’s it.
Not a word was said about treating tiredness, not a word about “cancer fatigue.”
There was no mystery.
We knew the gaps in our knowledge were vast, but doctors kept prescribing anyway, and profiting handsomely.

In 2004, the FDA called a meeting of its advisory committee to discuss Henke’s head-and-neck study and Leyland-Jones’s BEST.
The advisory group, called the Oncologic Drugs Advisory Committee, reviewed the data.
I was a member of that committee.
At the meeting, some of us vented, and our overarching recommendation was to urge the agency to monitor ongoing studies for further evidence of harm.

It’s important to remember what the FDA does and what it doesn’t do.
The agency approves the indications—the use of drugs.
It manages the package insert—the label.
It can tell you that a drug can be used to treat a specific disease.
It can make sure that advertised claims are consistent with the language of the label.
It can issue safety warnings, institute restrictions, revoke indications.

However, the agency is not permitted to regulate the practice of medicine.
If doctors want to use a drug in a way that goes beyond the indication on the label, they can do so unimpeded.
For example, the agency couldn’t regulate the use of high-dose chemotherapy and bone marrow transplantation for breast cancer.
This procedure made use of approved drugs.
That they were used in lethal doses in conjunction with removal and reintroduction of bone marrow fell under the rubric of the practice of medicine.
Similarly hemoglobin-building drugs, called erythropoiesis-stimulating agents (ESAs) were approved drugs.
They were approved on thin data because we needed an alternative to blood supply, which was at the time threatened by the AIDS epidemic.
At the time of approval, no one even suspected how widely these drugs would be used, or that someone would concoct the idea of pushing a cancer patient’s hemoglobin into the “normal” range, whatever that means.
Doctors were free to use these drugs as they saw fit even as tangible evidence of harm was starting to emerge.

Sure, the J&J direct-to-consumer advertising campaign promised too much.
Unfortunately, the armamentarium of penalties that the FDA can impose on companies that make unfounded claims doesn’t include the removal of indication.
The agency may send warning letters or, at worst, impose fines.
In the heyday of ESAs this wasn’t about to happen.
Warning letters were exceedingly rare, because in the Bush administration former pharmaceutical-company lawyers were running the enforcement arm of the FDA.

*

AN
uninitiated observer might have thought that misgivings about safety and tumor promotion would have depressed the sales of these drugs.
Just the opposite: gross sales of Aranesp and Procrit in oncology jumped from $3.783 billion in 2003 to $4.854 billion in 2006.
Most of this explosive growth was due to a spectacularly aggressive marketing campaign for Amgen’s Aranesp.

In a 2007 presentation for Wall Street analysts, an Amgen executive explained that the company sold two products: “the red juice” to fight anemia, and “the white juice” to fight neutropenia, a deficiency of white blood cells.
If you ran an oncology practice at that time, you needed to be able to buy both kinds of juice at the best possible price—and you were therefore completely beholden to Amgen.

This campaign went far beyond direct-to-consumer ads aimed at convincing patients to ask for treatment.
Amgen was offering doctors substantial rebates for bulk purchases, thereby inducing doctors to pump ever-larger amounts of Aranesp into each patient.

I obtained a copy of a 2006 supply agreement between Amgen and an oncology practice.
This is a proprietary document that doesn’t circulate widely, and you can see why.
It states that a practice that made a gross purchase of $422,800 worth of Aranesp in a single quarter stood to get back 18 percent of that amount—$76,100—in a rebate.
Practices that made larger purchases could get back up to 21 percent on the purchases they had made.
Also, Amgen offered discounts on its other drugs, Neulasta and Neupogen, so-called white juice.

These incentives not only induced practices to abandon Procrit in favor of Aranesp, but also made certain that the sales of hemoglobin and neutropenia drugs grew proportionally.
Amgen set its sales targets in dollar amounts of gross purchases, which a marketing expert such as Lilla would tell you is an inducement to increase the dose of the drug, giving more of it to each patient.
My colleagues in private practice tell me that they would receive calls from Amgen sales reps, informing them when they were underperforming, i.e., running short of earning the discount points.

Since I am an academic oncologist, my salary is not directly determined by the amount of services I sell.
I am paid to treat patients, but I am also paid to teach, conduct research, and publish.

Physicians in private practice—who treat more than three-quarters of cancer patients in the United States—are expected to generate certain revenues, and their take-home pay is usually determined by the amount of medical services and drugs they provide.

With these powerful incentives set in motion, many hospitals and oncology practices in the United States instructed nurses to ask leading questions about “fatigue” with the intent of expanding sales to a growing number of patients and upping the dosage to each patient.
In August 2007, during a conference call with investors, an Amgen official introduced a chilling term: an
ESA treatment opportunity.

What is an
ESA treament opportunity
?

A
treatment opportunity
was Amgen’s way of saying “a cancer patient.”
To increase their earnings, drug companies and doctors set out on a search for treatment opportunities, often forgetting about the sacred trust between doctors and patients.

*

WHILE
Lilla was experiencing a burning sensation under her skin, my colleagues were doing quite well, thank you.
A friend of mine suggested with bitter irony that oncologists with kids in college should be allowed to target a hemoglobin level of 13.5 g/dl.

Meanwhile, Lilla still had no reasons to ask questions about her treatment.
“I just went along with things and did what I had to do,” she recalls.
“I was feeling good, the treatment was working, and I was managing the chemo just fine.”

In 2006, Lilla met several women affiliated with Share, a New York–based suport group for women with breast and ovarian cancer.
Some advocacy groups do what doctors and drug companies tell them to do.
Share is different.
Share members are hard-nosed New Yorkers who seem to have an unlimited number of ways to make it amply clear that they don’t tolerate nonsense.

Initially, Lilla’s new friends asked her to help with peer counseling on the Share hotline.
She took calls from women whose experiences with metastatic breast cancer were similar to her own.

Lilla loved counseling.
“You pick up the phone, and somebody is just devastated because they have this horrendous diagnosis,” she recalls.
“You sit down and talk to them for a few minutes on the phone, and, for one thing, they hear that you are a survivor and that makes them feel good to talk to someone who has made it through.
And then they tell you about their emotional or medical concerns.
In most cases, if they have medical issues, I suggest a second opinion.
When I finished each day, I felt that I had gotten much more than I had given.”

Specifically, Lilla got a sense of how the US medical care system operated, not just in New York, but around the country.
For some reason, the stories that scared her the most originated in Florida.
She realized that these stories fit into broad categories.

There was the story about an uninsured or underinsured woman getting a breast-cancer diagnosis only to realize that she cannot get care.
The Centers for Disease Control and Prevention run a free screening program that pays for treating cancers diagnosed through screening.
Unfortunately, in most states the program runs out of money on the sixth or seventh month of the fiscal year.
CDC says it can pay for care for only 18 percent of eligible patients.