Marge Halford of Taylorville, Ill., holding the last of her Gleevec pills for the month. (Isaac Smith/For the Washington Post)

When the drug company Novartis launched its breakthrough cancer medicine, Gleevec, in 2001, the list price was $26,400 a year. The company’s chief executive acknowledged it was expensive, calling it an “uphill battle to win understanding for our decision.”

Today, that hill is a mountain. Since Gleevec was approved to treat a rare form of leukemia, similar drugs have come on the market — and the U.S. wholesale list price for a year’s supply of that little orange pill has soared to more than $120,000.

The pharmaceutical industry has insisted that the competitive market controls the costs of medications and that the overnight price hikes that have sparked public outrage and congressional investigations are outliers.

But Gleevec’s arc shows that even for a medicine that is the fruit of years of research — a prime example of what drug companies aspire to do — the market can fail. Instead of rising in sudden surges, Gleevec’s price crept inexplicably upward each year. When powerful second-generation drugs began to give physicians choices, Novartis raised the price even faster.

This price inflation helped turn Gleevec, a drug that was not supposed to make much money, into the biggest drug by revenue at one of the world’s largest drug companies.

“They say market forces set the prices reasonably, but there are no market forces,” said Hagop Kantarjian, chairman of the leukemia department at the University of Texas MD Anderson Cancer Center. “The drug companies are so few, they have carved out oligopolies.”

In a normal competitive market, prices influence what people buy — but not in health care. Brand-name drugs generally do not compete on price, because physicians and patients rarely pick treatments based on price — and often are not even aware what the prices are. Drugs each have a different benefit and side-effect profile, and doctors pick the drug they think will work best for their patients. What competition does take place occurs in secret negotiations between drugmakers and middlemen.

Which all points to a very strange fact about drug prices: They do not really exist. List prices are nothing more than a starting point for bargaining between drugmakers and the companies that provide prescription drug benefits. The cost for patients varies widely, influenced by discounts and rebates developed behind closed doors and applied in secret.

To track how Gleevec became a multibillion-dollar drug, The Washington Post used the median amount paid by privately insured patients and their health plans before discounts and rebates, an analysis prepared by health services researcher Stacie Dusetzina of the University of North Carolina at Chapel Hill using data from Truven Health Analytics.


Marge Halford at her home in Taylorville, Ill. (Isaac Smith/For the Washington Post)

One of those patients is Marge Halford, a 65-year-old nurse who lives in Taylorville, Ill., and has been taking Gleevec since 2009. The amount patients pay can vary widely depending on their insurance plan, and Halford’s cost started at $500 a month, but within a year the drug she needs to say alive was costing her more than $800. She and her husband considered divorce, hoping her single income was low enough to qualify for financial aid. But when they did the math, she still made too much money to get help.

About a year ago, sick of watching a whole paycheck disappear to pay for her pills every month and hoping to reduce the nausea and vomiting that are a side effect of the drug, Halford persuaded her doctor to put her on a cheaper, lower dose of Gleevec. Halford likes to say she is blessed — her kids are grown, her house is paid for and she has been able to find the money to pay for her medicine. But she is worried about retirement.

“The drug is so stinking expensive, and I don’t know what will happen,” Halford said. “The drug is a godsend. The price is not.”

A risky start

The odds were stacked against Gleevec from the beginning.

The disease it treats, chronic myeloid leukemia, afflicted a small number of people — about 4,500 new patients each year in the United States. If it worked — a big if in drug development — the numbers suggested it was unlikely to be a big moneymaker.

For major drug companies, a benchmark of success is a blockbuster drug that brings in at least $1 billion in sales each year. Scientists who worked on Gleevec’s early development recalled marketing projections that suggested the drug would peak at $100 million in annual sales.

“It looked pretty depressing,” said Nick Lydon, a scientist who headed the team that developed imatinib — the generic name for Gleevec — in the 1990s. Ciba-Geigy, the company he worked for, was not all that excited about the market for a rare leukemia treatment that used a risky new approach to attack cancer cells, but it decided to take the gamble.

Lydon teamed with Brian Druker, an oncologist and researcher at Oregon Health and Science University who tested the drug on bone marrow samples from his patients.

In 1996, Ciba-Geigy became Novartis in a merger. Two years later, Druker led the first test of the drug in people.


Brian Druker, director of the Knight Cancer Institute at Oregon Health and Science University. (Michael McDermott/PR Newswire)

The results were dramatic: Many patients experienced a massive reduction in the number of white blood cells, and in some cases cancer cells disappeared altogether.

Under pressure from Druker and patients, Novartis sped up development of imatinib, and in 2001 the drug earned the fastest U.S. cancer drug approval to that date.

That left the company wrestling with the delicate issue of price.

Novartis took “a huge financial risk by scaling up production to a multimillion-dollar level for a drug in early-stage development targeting a small market,” spokesman Eric Althoff said in an email.

In his 2003 book, “Magic Cancer Bullet: How a Tiny Orange Pill is Rewriting Medical History,” Novartis’s then-chief executive, Daniel Vasella, laid out the price considerations: the small patient population, the price of the existing treatment and the need to recoup the significant research and development costs of getting a drug to market. Based on those factors, the company settled on $2,200 a month.

“The result for Novartis: It would not stand to make a large financial gain,” Vasella wrote.

By the end of 2003, Gleevec was Novartis’s No. 2 drug, a billion-dollar blockbuster. Last year, it generated $4.7 billion in worldwide revenue, more than half of that from the United States.

The price rises

Novartis nudged Gleevec’s price higher slowly, at first. These were not the sudden, steep surges that have been an easy target for politicians but subtle increases that have gone unchecked throughout the industry.

The median amount paid by patients and insurers stayed stable for the first four years, according to Dusetzina’s analysis, hovering around $3,200 a month (in 2014 dollars adjusted for the inflation of medical products). Starting in 2005, the cost ticked upward gradually, at an average of 5 percent above inflation each year, to $3,757 a month in 2007.

Such incremental drug price hikes have become a defining part of pharmaceutical companies’ bottom lines, said Richard Evans, an analyst at SSR Health, an investment research firm. Evans likes to compare drug prices to a balloon without a tether. For the two decades preceding 2013, Evans estimates, annual price hikes of about 4 percent above inflation fueled pharmaceutical growth.

“That is unprecedented,” Evans said. “I can’t think of any other industry that has had that real pricing power. Structurally, that inflation has obviously been hugely important for the industry.”


Competition enters

In 2006, Bristol-Myers Squibb earned approval for a drug called Sprycel, or dasatinib, that would work in the same targeted way as Gleevec. Novartis developed a second-generation drug, too, called Tasigna, or nilotinib, that was approved in 2007.

As is typical, the new drugs entered the market above the price for the existing drug. The drugs were initially approved for a smaller patient population and offered a clear benefit over the existing treatment — people for whom Gleevec had failed now had another option. According to Dusetzina’s data, Gleevec’s median cost was $3,757 a month in 2007, compared with $5,477 for Sprycel and $6,929 for Tasigna.

The two drugs seemed to exert a magnetic pull on Gleevec’s price — upward.

According to Dusetzina’s analysis, the amount insurers and patients together paid for Gleevec accelerated right around the time its competitors were being introduced, as if it were playing catch-up. In 2008, the median cost jumped by 8 percent to $4,063 a month.

In 2010, Gleevec gained more direct competition from both drugs, which were approved for newly diagnosed leukemia patients. At this point, Gleevec’s price increases veered quickly into larger hikes that brought it closer to its competitors. An era of price increases of 10 percent or higher began.

Sales revenue at Novartis followed suit. In 2010, Gleevec’s annual global sales soared past $4 billion.

“What has been hard to justify, as competitor drugs have been developed, is they’ve entered the market at higher and higher prices and the price of imatinib has continued to go up to match them,” said Richard Larson, a hematologist at the University of Chicago. “Ordinarily, you might think with three equally effective drugs on the market, the price should go down through competition, but it’s been a failure of the competitive pricing process.”

Representatives of Novartis declined an interview request but answered questions by email. Althoff, the Novartis spokesman, said that the ability to raise prices is necessary to reflect not only changing market forces but also the evolving value of the treatment — how much it extends and improves the quality of patients’ lives.

He added that “price adjustments” allowed the company to take risks in research and development necessary to fuel innovation, particularly in rare cancers.

Testing a drug does require additional investment, but at the same time, companies developing drugs for rare diseases, also known as “orphan diseases,” can take advantage of federal tax credits.

The Orphan Drug Act offers tax credits for up to half of the cost of clinical testing for each designation — potentially worth tens of millions of dollars. Gleevec has seven orphan-drug designations, including the one for chronic myeloid leukemia.

Althoff would not respond to questions about whether Novartis had used those tax credits, saying the information was proprietary.


Rising demand

In 2001, the life expectancy for people with chronic myeloid leukemia was about five or six years. Today, their life spans approach normal.

“I joke, ‘We’re primary care doctors now,’” said Druker, the Oregon oncologist who played a key role in the development of the drug. “When my patients come in, I want to make sure they’re getting their mammograms, their colonoscopies, their cholesterol checked, their blood pressure — because it’s as likely they’re going to die of something else.”

That has meant a steady rise in the number of patients living with the disease and taking Gleevec or its competitors. A 2012 study published in the journal Cancer found that, before Gleevec came along, the estimated prevalence of the disease in the United States was between 25,000 and 30,000 people. Because of the drug’s success, that number is projected to have tripled already and to reach 112,000 people in 2020. At the same time, the drug has earned additional approvals for other rare cancers.

“If the expectation was those . . . patients only take the drug for a year or two before it lost its effectiveness, then it seemed reasonable to allow the drug company to profit through its development and marketing,” Larson said. “But in fact most patients do have durable responses. They stay on the drug essentially lifelong.”

In a patient newsletter in 2001, Vasella said the small patient population was a key factor that required such a high price and that his company “might be able to lower the price” if more patients began using Gleevec for other cancers.

Although Gleevec never became a drug for a large number of patients, it surpassed expectations and was approved for other rare diseases.

Despite that, the price continued to rise.


Access

A lifesaving drug that patients cannot access because it is too expensive would be a public relations nightmare, and pharmaceutical companies take steps to make sure that does not happen.

Althoff noted that the majority of leukemia patients pay no more than $100 a month for their pills. Novartis has provided the drug free or at reduced cost to an average of 5,000 people in the United States each year for the past 6 1/2 years.

According to Dusetzina’s analysis, the median co-pay for privately insured patients has barely budged in the lifetime of the drug, rising from $16 per month in 2001 to $33 in 2014. At the same time, the total amount paid for the drug has risen from a median of $3,271 per month in 2001 to $8,156 in 2014. Those figures do not take into account discounts or rebates but suggest that insurance is picking up a large amount of the tab.

The rest of that cost is spread across the health-care system, through premiums and deductibles.

Developing innovative drugs costs money, and this may be society’s solution to paying for it. But even a relatively small co-pay can affect people’s health.

Dusetzina’s research, published in the Journal of Clinical Oncology in 2013, found that higher co-pays affect whether people keep taking drugs. In her study of chronic myeloid leukemia patients, nearly 1 in 5 with co-pays above $53 a month discontinued their drugs in the first six months.

And while Dusetzina’s data on co-pays gives a glimpse of how much people with private insurance pay, it does not reflect everyone. According to Medicare data released late last year, total spending on Gleevec rose from less than $400 million in 2010 to nearly $1 billion in 2014, not including privately negotiated discounts. People on Medicare’s Part D prescription drug plan who do not receive a low-income subsidy pay $525 a month on average, according to a 2010 Government Accountability Office analysis.

That’s Dianne Dale Watson’s experience. The 77-year-old retired psychotherapist from Eugene, Ore., said she has saved throughout her life, “like something was chasing me.” For the past nine years, she has found herself doling out those savings, $500 a month, to pay for Gleevec. The drug saps her travel budget to visit her grandchildren in Madison, Wis. She has even made up a song about her prescription drug costs.

“The cost of drugs is a high cost. Where it stops, nobody knows,” Watson croons in a high, clear voice. “These drugs that I take are not optional. They help me to go on my way. But Medicare D is dysfunctional, best wishes and have a nice day.”


Going generic

Gleevec’s long, game-changing ride is nearing a new chapter in its story. Its patent exclusivity in the United States ended last month, opening the door for generic competition.

Andrew Hill, a senior research fellow at University of Liverpool, has analyzed how much it costs to make the drug, imatinib, in raw ingredients. A year’s worth of drug, made into tablets and bottled, with a 50 percent profit factored in, would cost no more than $216.

“We have to take away the sort of mystique about this,” Hill said. “They’re just chemicals.”

A patent litigation settlement between Novartis and Sun Pharmaceutical Industries, the first company to produce generic imatinib for the United States, delayed the generic’s launch by seven months. Now, Sun Pharma says its price for generic imatinib is 30 to 50 percent less than Gleevec’s list price. The company will have six months of exclusivity, and after that the door will be open for other generic competition. The price is expected to fall by 70 to 90 percent off of the brand-name price in the first year, according to University of Chicago health economist Rena Conti.

But Gleevec’s last act has been a profitable one. Novartis has hiked Gleevec’s price more rapidly in recent years — including a whopping 19 percent increase between 2013 and 2014, from a median of $6,841 a month to $8,156, according to Dusetzina’s analysis.

“You could replace it with a mental-health or cardiac drug; it would be exactly the same story,” Conti said. “They always raise the branded price right before entry.”

Novartis’s Althoff noted that with discounts, Gleevec is cheaper than other treatments for chronic myeloid leukemia.

Several oncologists said there appears to be a subtle effort by pharmaceutical companies to steer physicians to the next-generation drugs, although there is not yet evidence that they help people live longer. In a 2014 news release, Novartis said that Tasigna, its second-generation drug, had “higher rates of early, deep and sustained molecular responses.”

But Vinay Prasad — an oncologist at Oregon Health and Science University who has been critical of cancer drugs that have been approved for stopping cancers from progressing but not saving lives — wonders what that means for patients.

“The question is, ‘Does it have anything to do with anything in your life?’ ” Prasad said.

Halford, the Illinois woman who has for years prayed for a generic, last year found that, after years of trying to get various kinds of financial assistance, something changed. She switched pharmacies and, to her surprise, Novartis offered her a discount. Suddenly she owed only $10 a month, a steep discount from the $800 she once paid. Watson recently saw her co-pay fall to zero — for reasons she does not understand but does not want to jinx.


Marge Halford's Gleevec pills. (Isaac Smith/For the Washington Post)

“It was wonderful, but why all of a sudden after six years are you giving me the co-pay?” Halford said.

She does not know why the switch happened, but she can’t help but wonder whether it is a ploy to prevent her from switching to a generic.

When Sun Pharma launched its generic, it announced its own $10 co-pay program. Novartis’s competition, it seems, has arrived.