If you want a glimpse into the future of medicine, look at Mary Sullivan.

The 63-year-old retiree from Corning lives anxiously with lung cancer, a disease that causes more deaths each year than breast, prostate and colon cancer combined.

Progress in treatment has been slow, but there are encouraging signs.

Her doctors at Roswell Park Cancer Institute compared her tumor cells with her healthy cells in search of a gene mutation that is found in about 15 percent of patients in the United States and that can cause uncontrolled cell growth. Knowing that she carried the mutation, doctors put her on the drug Tarceva, which targets the mutation with fewer side effects than standard systemwide chemotherapy.

“The drug is working. I’m still here and am content with that,” said Sullivan, who was diagnosed with advanced disease in the fall of 2010.

The importance of Sullivan’s story is that it’s now believed that cancers develop from mutations in our genes. Some mutations are inherited from parents, but most are acquired from other factors, including such environmental exposures as cigarette smoking.

The thinking goes that if we can decipher patients’ genes, better tests and treatments can be targeted to their genetic abnormalities, such as in Sullivan’s case. It’s called personalized medicine, and scientists in Buffalo and elsewhere around the world are betting on it in a big way – for cancer, as well as for other diseases.

“Personalized medicine will change how we view medicine and oncology. You are going to see cancers classified by their mutations instead of the organ where they originated,” said Dr. Candace Johnson, deputy director of Roswell Park.

The Western New York Regional Economic Development Council, one of 10 regional councils appointed last year by Gov. Andrew M. Cuomo to spur economic development, awarded one of its largest grants – $5.1 million – for a new Center for Personalized Medicine at Roswell Park, and it is now getting under way. The cancer center invested $16 million, much of it for a highly secure supercomputer and machines called sequencers to analyze patients’ genes.

Like a small-market sports team that must find a strategy to compete against bigger and better-funded rivals, researchers here are investing initially in three initiatives that build on existing medical strengths:

• Breast cancer: Recent studies have identified different subtypes of breast cancer, meaning that many patients receive the same chemotherapy even though their tumors are different and may not respond well to a treatment that also comes with serious side effects. One project will search for a way to determine which of two main types of standard chemotherapy works best.

• Bio-banking: With the aid of a mobile unit, an effort will soon begin to obtain blood samples, and demographic and health information from hundreds of individuals representing the ethnic, racial, socioeconomic and geographic diversity of the region. The idea is to start creating a genetic profile of the community that can identify health care priorities and help design trials of new tests and treatments.

• Bladder cancer: In collaboration with Western New York Urology Associates, a large specialty group, another project will focus on developing a test based on genetic variations for earlier detection of superficial bladder cancer, the ninth most common cancer and among the most expensive cancers to treat per patient.

Roswell Park also is working closely with Computer Task Group, the Buffalo-based company that specializes in the installation of electronic medical record systems in hospitals. The organizations intend to market a service that combines the institute’s ability to analyze the genetic differences in cancer patients with CTG’s expertise in medical records.

Complex links

If the collaboration works as planned, physicians would send patient tumors to Buffalo for genetic analysis, and the information would be incorporated into the patient’s electronic medical record in a way that helps physicians make treatment decisions.

“Too often in cancer, time and money are wasted using the incorrect drug,” said Michael Colson, senior vice president of information technology solutions at CTG. “If we can make this information available, then doctors can target the best drug at the right time and at the right dose.”

Personalized medicine remains more promise than reality. It’s still very expensive to decipher an individual’s genetic profile, and the technology to do so is still limited to research centers.

Scientists also are finding that the links between gene mutations and disease are complex and don’t necessarily predict how things will turn out. Many conditions develop in the cells for a combination of reasons.

In addition, personalized medicine is provoking difficult questions.

Will insurance companies pay for widespread genetic testing of patients when it’s not yet entirely clear how or whether it improves health outcomes?

Can sensitive genetic information be stored and managed in a way that safeguards patient privacy?

Will companies rush to commercialize tests for genetic risks without clear standards for accuracy and advertising?

How do we ensure that the benefits don’t go to only those who can afford it?

“The science is moving fast, as is our ability to collect genetic information and process it,” said Dr. Arthur Caplan, head of the division of medical ethics at New York University’s Langone Medical Center. “What isn’t moving so fast is the legal and regulatory environment over the way genetic information is collected, stored, used and sold.”

All of this seemed like science fiction a few years ago because of the complexity of our genetic blueprint.

A twisted ladder

Every cell in our body contains DNA, the genetic instructions passed down from one generation to the next needed to make and maintain an organism. DNA looks like a long twisted ladder, with the steps composed of pairs of four building-block chemicals. There are about 3 billion of those steps all arranged in a certain order.

The DNA of every person in the world is more than 99 percent identical. It’s that remaining part that makes us different from each other, including our risk of disease. Cancer can occur when mistakes or changes occur in the DNA sequence.

A gene is a section of DNA that tells cells to produce proteins that are responsible for just about all of the work our cells do, determining everything from what we look like to how well we respond to an infection. Humans have 20,000 to 25,000 genes, and every person has two copies of each gene. The complete set of those genes is called a genome.

One of the key reasons why there is so much excitement about personalized medicine is that the time and cost of figuring out the exact order of an individual’s genome – a process known as gene sequencing – is declining to levels where it could become a routine part of medical care.

Gene sequencing

The first human genome was sequenced in 2003 after more than 13 years of work at a cost of nearly $3 billion.

Today, technical advances allow gene sequencing machines to decipher an individual’s complete genome in a matter of days for about $10,000, and experts expect the price to continue dropping.

It’s conceivable that patients in the future will have their genomes sequenced to find out what diseases they are at risk for and work with physicians on prevention strategies. But, with the astronomical cost of health care, it’s also conceivable that questions will arise over which patients truly need gene sequencing and, if so, when.

In cancer, another goal is to take DNA from a patient’s normal cells and compare it to the DNA in a tumor cell, looking for differences that identify a mutation in a gene and then tailoring the treatment to that particular gene mutation.

“Every cancer is probably 20 different subtypes, depending on the irregularities in genes. Gene sequencing could be transformative,” said Dr. Carl Morrison, executive director of Roswell Park’s Center for Personalized Medicine.

Managing data

Roswell Park officials are talking about soon making gene sequencing standard for patients with lung, skin and colorectal cancers, with plans for adding other cancers as the science advances.

Aside from costs, managing the amount of data from gene sequencing poses a challenge.

There are millions of potential genetic differences in a human genome and, like needles in a haystack, only some play a role in a particular disease.

On the horizon

Sullivan’s gene-targeting therapy for offers a preview of what’s on the horizon.

An analysis of her tumor identified a mutation in a gene known as EGFR. The mutated gene produces a protein called epidermal growth factor receptor, and the drug Sullivan takes blocks the protein’s action.

The drug treatment is far from a cure. But it appears to markedly improve survival for thousands of patients.

And for someone who was initially given no more than six months to live, each day of life is like a gift.

“You fret about what might be,” Sullivan said. “But you have to have faith in the doctors and the treatment.”