Diagnostics & Imaging Week Washington Editor

WASHINGTON — Personalized medicine (PM) may be a mantra for many healthcare researchers, but Friday's session on the topic, hosted by the Food & Drug Law Institute (FDLI; Washington) and the American Academy for the Advancement of Science (AAAS; Washington), made quite clear that there are far more questions than answers facing the concept.

Perhaps the most pressing of these is the question of where modern medicine will find all the money needed to move personalized medicine from a Star Trek future to a mundane present with the help of diagnostics.

Susan Dentzer, editor-in-chief of the healthcare policy publication Health Affairs, said reports of the decoding of the human genome led to stories that genes "were like tarot cards, but for real" and could provide a "very clean and very accurate" picture of one's health risks. The news also fed the perception that "we were in an immediate era of personalized medicine, and all this would take care of itself," she said.

"We all know that one of the fruits of genomic and proteomic medicine is the biomarker," Dentzer said in outlining a hypothetical situation to highlight some of the issues surrounding PM. This hypothetical biomarker, dubbed Luc-3, "could be detected in blood tests."

Along for the ride with Luc-3 was a hypothetical physician named Samantha Anderson, MD, who can make use of Luc-3 to evaluate her patients for breast cancer.

Dentzer quickly delineated one problem. "Some doctors and researchers immediately jump in and advocate immediate testing of all women for Luc-3," she said, but she pointed out that Dr. Anderson has to figure out how to counsel her patients in the face of vastly different circumstances of each patient's life.

One of her patients, Terri, is a 30-something wife and mother with a family history (her mom and aunt both died of the disease in their 50s), and the family has no health insurance, while Leslie, 45, "has terrific employer-sponsored health insurance." Dentzler said Leslie is not sure she wants to tell her boss because of the possibility of missing out on an opportunity for advancement.

Anderson also treats Wanda, a 60-year-old African-American woman, as well as Suzanne, 75, a Medicare beneficiary who also has hypertension and diabetes.

In this scenario, Luc-3's presence in cases in which the cancer has spread to lymph nodes is associated with a 95% probability of cancer and a 79% rate of post-surgical recurrence, assuming medium to large tumors. For women who differ from this first group only by exhibiting no lymphatic invasion, 80% would exhibit Luc-3 presence and as a group would experience a 47% probability of recurrence after surgical resection. Only 20% of the women with small tumors would return Luc-3 positive assays, and their rate of recurrence is pegged at 10%.

The question was which of these patients should undergo post-surgical chemotherapy, and Dentzer asked a panel of nine experts, "what factors should Dr. Anderson consider when advising her patients?"

Howard Levy, PhD, a geneticist with Johns Hopkins Hospital (Baltimore), addressed "the fallacy of genetic determinism."

Genetic make-up, Levy asserted, "is not the end-all, be-all," so Luc-3 "is not going to give the final answer" to the question of how these patients will fare on chemotherapy. He said he would advise Anderson to "act as an adviser" as much as a prescriber. Levy also made the case that "if we as medical professionals want to order a test, we need to know what we want to do with that information."

Annette Bar-Cohen, of the National Breast Cancer Coalition (Washington), said given that diagnostics are still somewhat primitive, "we have time to get educated" on what the biomarker says about outcomes for each of the four women. She also pointed out that doctors, not just patients, need education, "because these days, the information is so complex" that even doctors have a hard time keeping up.

Whatever the chosen course of therapy, "in almost all these cases ... I would recommend that they enroll in a clinical trial" to bolster data that will help avoid inappropriate treatment and non-treatment, Bar-Cohen said.

Finley Austin, PhD, of Hoffman-La Roche (Nutley, New Jersey), said, "we have to distinguish if Luc-3 is a prognostic marker," but "my advice to almost all these women is ... this should be done in the context of a clinical trial.

"What's key to look at is what is the level of evidence" for Luc-3, Austin said, because payers are leery of coughing up for unproven diagnostics and treatments.

Gregory Downing, PhD, program director for the Office of Personal Healthcare at the Department of Health and Human Services, posed the question of how to fund all this. He said he would tell Anderson that "there are many scenarios like this emerging," and that she might want to establish the rate of the biomarker's presence in healthy women and question whether it is a legitimate surrogate for breast cancer outcomes.

"Recent studies have indicated that lowering lipid levels" does not have the expected impact on heart disease, he pointed out.

"I think what we're talking about is establishing an evidence base," Downing said, adding that "our goal should be looking at ways of improving evidence development" and giving a boost to the federal data collection infrastructure.

"Should this take precedence over pandemic planning?" he asked, referring to other HHS imperatives, such as FDA funding.

Privacy issues also arose during the discussion. Among the panelists was Carol Barash, PhD, the principal at Genetics, Ethics and Policy Consulting (Boston), who said, "There is the possibility that the information" on a clinical trial patient "may lead to privacy violations in terms of data mining."

Dentzler remarked that most patients might say "right now, I'm more concerned about living than I am about privacy."