Personalized cancer therapy is no longer just an exciting prospect, and better survival rates - as well as escalating spending - are proving it.

"It's here. It's definitely here," said Pasi Jänne, an oncologist at the Dana-Farber Cancer Institute in Boston. "Today, it's the most effective way to treat patients: figure out the genetic fingerprint of an individual's cancer and tailor the therapies to it."

This year, President Obama announced an initiative focused on "precision medicine." But in oncology, the model has arrived. Targeted therapies such as Avastin and Gleevec that disrupt the activity of genetic mutations have dramatically increased their share of global oncology sales. In the U.S. and other developed countries, that share has risen from 11 percent in 2003 to 48 percent last year, according to a new report by IMS Health, the health-care market-research company with a division in Plymouth Meeting.

Though molecularly targeted drugs are not yet available for most cancer patients, they account for the majority of the 45 new cancer drugs launched since 2010, the report found. And because the same mutation may drive cancer that arises in different organs, targeted drugs often get approved to treat multiple cancers. The newest innovations, which activate the immune system, work even against metastatic malignancies, notably lung cancer and melanoma, that are unstoppable with conventional surgery, chemotherapy, and radiation.

But the report also identifies downsides. Personalizing therapy increasingly requires molecular diagnostic tests to identify patients with the relevant target. Precision therapies, which generally have fewer toxic side effects than chemotherapy, are often taken indefinitely, or until they stop working. That is adding to the complexity and costs of treating cancer, posing dilemmas for oncologists, health insurers, and governments.

Targeted drugs "do bring with them improvements in overall survival," said IMS Health senior vice president Murray Aitkin, an author of the report. "The broader issue is how much are we willing to invest in winning the war on cancer - and who is paying?"

In the U.S., the report found, two-thirds of Americans diagnosed with cancer now live at least five years, compared to just over half in 1990. The U.S. is still the world leader in oncology spending per capita - about $100 per person per year - partly because drug prices are higher here than in other developed countries. The average cost of new oncology drugs is now $10,000 a month, up from $5,000 a month a decade ago.

In the near future, the report noted, treatment is likely to involve combinations or sequences of drugs aimed at a bunch of molecular targets, possibly with immunotherapy.

For an example of the evolving complexity, consider the treatment of metastatic lung cancer. A subset of patients - about 15 percent of whites and 40 percent of Asians - have a mutation in EGFR, a gene that, when healthy, allows cells to grow and divide. The mutation can be targeted by EGFR inhibitors including Tarceva (erlotinib) and Iressa (gefitinib), but most patients' cancers soon become resistant to the drugs, usually by developing a new mutation known as T790M.

The second-generation EGFR inhibitor Gilotrif (afatinib) is not very effective against this mutation, so third-generation EGRF inhibitors that target T790M are in development.

Last month, studies of two of these third-generation drugs showed that about 60 percent of patients with the T790M mutation responded. One drug staved off disease progression in half of responders for at least 10 months; the other drug, for 13 months.

Jänne, who led one of the studies, said quantifying the benefits was also complex.

"It's more than just looking at the costs of the medicines. It's more what patients are able to do during treatment - things they can't do with toxic chemotherapies. Patients can take these chronically and go back to work. Our challenge and opportunity is to figure out in whom are the drugs most effective, and for the longest period of time."

Still, the report predicts that as health insurers confront high per-patient costs, they may come up with restrictions, perhaps denying coverage of a new drug that doesn't extend survival by a certain number of months.

"We see that being done by payers in other countries, where new drugs are subject to some mathematical calculus of where they provide value," Aitken said. "In the U.S., particularly with oncology, we've seen a limited appetite for that."

Beyond costs, personalized medicine is putting new demands on oncologists.

"We're on the edge of a real explosion in the number of options, and combinations of options," Aitken said. "Most oncologists are not in big academic medical centers. It's a challenge for them to have the sorts of tools and support they need."

215-854-2720 @repopter