American healthcare is expensive, and the political debate continues to rage about how best to pay for it and reduce costs, particularly for Obamacare. One sign of the financial stress: “Medical bills have become the single-largest cause of personal bankruptcy,” notes a recent article in the Journal of the American Medical Association (JAMA).

But as health care continues to eat an increasingly bigger chunk of the U.S. GDP, while not scoring particularly well on international quality of care measures for advanced countries, there is another strong undercurrent in medicine that is advancing in a positive direction. That is: a revolutionary approach to tackling disease called precision medicine (or personalized, or genomic medicine), which promises to deliver care that is both more lifesaving and more cost-effective than current therapies.

While definitions vary (often a symptom of a new field), the basic principle is to identify disease-causing abnormalities in a patient’s genetic profile — or genome — and then to develop more highly targeted drugs to attack those abnormalities. Patients that don’t have these mutations can be spared the pain and expense of treatments that won’t work for them. The field is about 15 years old — young in the annals of medical innovation.

Just how transformative precision medicine can be is reflected in the story of the cancer drug Gleevec (imatinib) for chronic myeloid leukemia (CML), a rare, fatal blood cancer that affects about 8,000 Americans a year. Thanks to this drug, introduced in 2001, the disease today is generally treatable.

Justin Bekelman, a professor of radiation oncology, and medical ethics and health policy at the University of Pennsylvania explains that most cases of CML are caused by a genetic abnormality called the Philadelphia chromosome, discovered by researchers at Penn and Fox Chase Cancer Center in 1960. Gleevec targets the abnormality in the simple form of a pill taken once or twice daily.

“It works better than chemotherapy and causes fewer side effects,” says Bekelman. “After over a decade, about 4 in 5 patients who take imatinib for CML had a complete cytogenic response, which means the cancer went away and stayed away. So that’s really a success [that] changed how we view cancer and its treatments.”

“The promise of precision oncology is to create a future — as many have imagined — where treatments are highly specific, minimally toxic, and dramatically effective.” –Justin Bekelman

Steven Joffe, a professor of pediatrics, medical ethics and health policy at Penn, characterizes Gleevec as “a medical grand slam.”

There have been a handful of other precision medicine wins as well (although none as dramatic as Gleevec), mostly to treat certain types of lung, skin and breast cancers. Cancer care is the area where precision medicine is making the greatest strides. Oncologists look at the genetic profile of both the patient and their tumor, since tumors acquire genetic changes in the process of becoming cancerous.

Yet today, says Bekelman, “the majority of cancers are [still] treated with chemotherapy that lacks specificity to the cancer.” While these traditional treatments can be lifesaving, they may have much-dreaded side effects: extreme fatigue, severe nausea, vomiting, pain and hair loss. And cancer recurrence is common. The promise of precision oncology, he says, is “to create a future — as many have imagined — where treatments are highly specific, minimally toxic and dramatically effective. And the question is whether that’s really promise, or hype.”

Lackluster Performance

Bekelman isn’t the only one to raise the question of hype. A recent article in the New England Journal of Medicine noted, “The widespread dissemination of the idea of personalized oncology has spread faster than the underlying science.”

The article discussed the latest precision cancer medicine approach, in which analyzing a tumor in an individual patient enables the selection of drugs to effectively control it. While this idea is “appealing to patients and foundations that support cancer research, and the molecular analysis of tumors is being marketed directly to patients,” there is a lack of evidence to support its benefits, the article said.

“Technology and marketing are way out in front of the science and the policy,” agrees Janet Weiner, an associate director of health policy at the Leonard Davis Institute of Health Economics of the University of Pennsylvania, in her blog on the institute’s website.

Some “99.99% of the cancer cells in a patient’s cancer might be effectively treated or killed by [a particular] drug, but .01% might have a mutation … that makes them resistant.” –Steven Joffe

Weiner writes that the percentage of patients receiving targeted (precision medicine) cancer therapies increased from 11% in 2001 to 42% in 2011. Private insurance payments for such therapies grew from 22% to 63%. Yet there has been only one randomized, controlled clinical trial of the precision medicine approach, a study called SHIVA. SHIVA compared a genetically targeted strategy to standard care in over 700 patients with a variety of tumors. No significant difference was found in disease progression or mortality.

Other, more limited studies have been “just as discouraging,” according to Weiner. And one precision oncology drug, Avastin, had one of its indications withdrawn in 2011 (for use in advanced breast cancer patients). The FDA determined that the drug had life-threatening side effects and was not helping breast cancer patients live longer or significantly control their tumors.

Why Haven’t We Cured Cancer?

Why don’t we have more Gleevecs? Why hasn’t precision medicine come up with drugs to control every patient’s cancer?

The challenge lies in the nature of cancer tumors themselves, says Joffe. Whereas just one genetic change causes CML, and four or five changes cause a certain type of lung cancer for example, most tumors contain a sprawling network of genetic changes. A biopsy may not even reveal changes occurring in other parts of the same tumor because genetic changes can vary within the same tumor lesion, or across different sites of metastatic disease. Bekelman points out: “Even though the promise of precision cancer medicine is so appealing, some argue that cancers are too complex,” with countless mutations.

Another problem, Joffe notes, is that “cancer is smart.” Cancer cells can develop resistance just as certain bacteria can become stronger when exposed to antibiotics. While “99.99% of the cancer cells in a patient’s cancer might be effectively treated or killed by [a particular] drug …  0.01% might have a mutation or acquire a mutation that makes them resistant.” The tiny remaining portion survives and grows, and the original medication no longer works.

Joffe distinguishes the “parsimonious” approach of identifying the few well-known, actionable tumor changes and prescribing accordingly from the more recent idea of sequencing an entire tumor and trying to figure out which of dozens or hundreds of mutations can be effectively targeted. “That broad strategy, which people are very excited about… has not yet been proven to improve outcomes,” he says.

On the same note, Bekelman cites a key study in which breast cancer patients were screened to receive precision oncology therapies specifically targeted to their genetic profile. Only 3% of patients screened had even a partial response to therapy, or a stable disease level. “Yes, that’s potentially meaningful for those patients,” says Bekelman. “But from a big-picture perspective, it’s really only a modest response… we don’t even know, maybe those patients would’ve had a response to something else.”

He adds, “the promise of precision medicine was to make really big, profound improvements in care for many people. That promise has become a reality in only a few uncommon examples.”

Medicine at $100,000 a Year

On top of the current lackluster showing of the broad precision oncology approach, there are the high costs. The new precision medicine cancer drugs win the prize for being astronomically expensive, routinely exceeding $100,000 a year. Another article in JAMA, which cites this price tag also asserts that cancer drug prices are rising faster than those in other health care sectors. What’s more, the pricing models are “not rational but simply reflect what the market will bear.” Weiner notes in her blog that from 1995 to 2013, launch prices for cancer drugs increased by 10% annually with no corresponding increase in effectiveness.

How can patients, payers and society continue to support these costs? Weiner points out that the very model underlying precision medicine is likely to make the situation even worse. “If these [precision] cancer drugs are better targeted, meaning they have a smaller population in which they work, then to recoup investments and get a good return on investment you can imagine that the price would have to be higher — because you’re treating fewer patients.”

Joffe agrees that precision medicine “may well increase” the economic burden of health care because the drugs are so expensive. Weiner says, “There needs to be some kind of business model for a drug that is effective in fewer people.”

The pricing models are “not rational but simply reflect what the market will bear.” –Janet Weiner

She notes that the U.K. prices drugs based on value. NICE, Britain’s National Institute for Health and Care Excellence, applies a threshold of around $33,000-$49,000 for each additional year of good health when deciding which treatments should be available on the National Health Service, according to The Economist recently. And some European countries handle their drug launches as “managed entries.” The price is set for the first year but then is re-evaluated based on whether the medication has proven more valuable than existing therapies.

Bekelman and Joffe are co-chairing a multi-disciplinary group at Penn, the Gant Family Precision Cancer Medicine Consortium, which is trying to unravel some of the knotty issues around precision cancer medicine’s cost as well as effectiveness. The group hopes to publish a policy framework that will support the success of precision cancer medicine going forward. Weiner, also a consortium member, says that the team includes clinicians, payers, drug companies and policy makers.

“We’re trying to develop policies that try to ‘pay more for drugs that do more,’ and also understand the value of cancer drugs,” says Weiner of the consortium’s work. “What framework do you use, and how do you get everyone’s perspective to decide on a reasonable price?” These decisions, of course, have profound reverberations for millions of people. “It’s cancer. Cancer is different. You’re talking about lives at stake.”

Bekelman says he still believes in the promise of precision cancer medicine. However, he says “we need to be clear-eyed and data-driven … as we make these very important investments in the future.”