Kellie Carey's doctor finally stopped dodging questions about how long she had to live six weeks after he diagnosed her lung cancer.

'Maybe three months,' he told her in his office one sunny May morning in 2010, she recalls.

Yet she is still alive, a testament to the most extraordinarydecade of progress ever in the long scientific struggle against lung cancer.

Tests found Ms. Carey's lung cancer to be of a rare type that researchers had found just three years earlier by deciphering its genetic code. The 45-year-old businesswoman in 2010 went on a drug Pfizer Inc. was testing for that type. By pinpointing her cancer, the drug probably helped give her years more to live than chemotherapy would have, her doctors say.

That is remarkable because lung cancer for decades defied efforts to find drugs that could extend an average patient's life by even a few weeks.

But an explosion in knowledge about the genetic mutations that cause tumors is just now offering the first real promise of drugs that can control what is the most-common and most-deadly cancer.

Ms. Carey has one of at least 15 lung-cancer variations, almost all of which scientists didn't know existed 10 years ago. Researchers have identified those variations, most of them in just the past four years, by decoding DNA in tumors--akin to how crime labs analyze DNA to genetically fingerprint suspects.

The newfound variants have led major cancer centers to revamp their approach to treating cancer and have spurred a rush among drug companies to find medicines that narrowly target each one.

The drugs don't cure cancer and face significant hurdles. But doctors now talk of a 'precision medicine' approach in which those pinpoint drugs can treat tumors far more effectively than catchall chemotherapy.

'What we're seeing is the beginning of a revolution in therapeutics,' says Janet Woodcock, director of the Food and Drug Administration's Center for Drug Evaluation and Research. 'We can only hope that this gets us to where cancer is managed or curable.'

Among signs that revolution really is afoot: A June 2013 study found that lung-cancer patients who were treated with drugs targeted at their genetically identified varieties lived 1.4 years longer than patients on chemotherapy whose cancers weren't genetically identified.

In effect, lung cancer is no longer a few common diagnoses. Instead, it is a growing list of rare cancers, each a target for its own drug regimen.

'It's likely that more than half of tumors have some alteration we can target with a drug,' says John V. Heymach, a lung-cancer specialist at MD Anderson Cancer Center, Houston. 'They may not all have the same success, but we know that in many cases, a targeted agent will work very well.'

The same goes for other malignancies: Scientists have decoded tumor DNA from breast, colon, kidney, skin and other cancers in recent years to discover scores of variations they didn't know existed before.

Research hospitals like MD Anderson, Vanderbilt University and Massachusetts General Hospital are among a growing number of cancer practices that routinely decode the tumor DNA of most patients with advanced cancer.

The lists of newfound variations have invigorated the drug industry, with companies like Pfizer, Roche Holding AG and Merck & Co. racing to develop drugs that target each one.

Last year, nearly 1,000 cancer drugs were in clinical development, up 52% from 2006, says the Pharmaceutical Research and Manufacturers of America, a trade group. The 'vast majority' of that growth is from drugs targeted at genetic mutations, says Bill Chin, the group's head of science and regulatory affairs.

Three drugs are on the market for newly discovered lung-cancer mutations. Dozens more are in clinical trials. Some approved for other cancers appear effective for specific lung cancers. And drug companies are targeting other mutations of all cancer types.

At least half the 27 medicines on Novartis AG's current list of oncology drugs in clinical development target cancer mutations. Precision medicine is 'fundamentally changing the way we think about cancer drug development,' says Hervé Hoppenot, president of the company's Novartis Oncology unit.

Just last year, the FDA established a 'breakthrough therapy designation' to hastenapproval of experimental drugs that show striking benefits in early trials, including those targeted at cancer mutations.