Genome Project's Likely Impact On Cancer Care: Limited In Short Term, Unpredictable In Long Term

In "Gazing into a crystal ball -- Cancer therapy in the post-genomic era," Mark Ratain, M.D., of the University of Chicago, and Mary Relling, Pharm.D., of St Jude's Children's Research Center, the chair and vice-chair of the Pharmacogenetics of Anticancer Agents Research Group, caution that before genetic information allows doctors to tailor therapy to each patient, much more work must be completed.

"Although access to a patient's genetic information will someday make cancer therapy more effective," note the authors, many "technical, experimental and clinical advances must be made before this day can arrive."

"The basic process," they add, "of target identification, drug discovery, preclinical and clinical work will be facilitated -- but not necessarily revolutionized -- by the human genome sequence."

For guidance in predicting the future, Ratain and Relling turn to five basic principles of prognostication, as formulated by technology columnist Robert X. Cringely, host of the PBS-TV miniseries "Triumph of the Nerds." The five principles are:

-- 1) We tend to overestimate the amount of change that will take place in the short term.

-- 2) We tend to underestimate the amount of change that will take place in the long term.

-- 3) The more specific a prediction, the less likely it is to be correct.

-- 4) Past performance is a predictor of future results, but not a good one.

-- 5) The most reliable predictions are those that follow established trends.

The cost of individualized therapy appears to confirm principle one. While the genome project has reported the sequence of a few volunteers, individualized therapy will require personal sequence information from each patient. At the current rate, that would cost about $120 million per person. If sequencing costs continue to fall by about 50 percent a year, in line with rule five, the cost would become reasonable, at around $500, in 18 years.

Their sobering point is that despite this remarkable achievement, and even if we are able to follow it by mapping out all the genetic changes that play a role in cancer, a daunting amount of basic biology and pharmacology will remain to be tackled.

Consequently, cancer research is about to enter a transitional period, in which there will be "an abundance of genomic data," note the authors, "but an inability to use it to fully predict cancer treatment outcomes."

One sterling example is the recent success of STI-571, an unusually potent medication that targets a crucial gene defect in one particular type of leukemia. That gene defect, a translocation of chromosome 9 and 22, was first described by Janet Rowley of the University of Chicago -- in 1973, 28 years ago.

"Many of the root causes of cancer lie in genetic abnormalities," note the authors. Publication of the human genome project is a necessary first step in optimizing treatment of cancer and other human diseases. It will help identify the molecular differences between normal tissue and tumors, speed the identification of new treatment targets, improve classification of tumors and facilitate individual drug dosing -- which is the goal of the Pharmacogenetics Group.

Rule two -- we tend to underestimate long-term change -- applies to those who say this day will never come. But the more pessimistic rules one, four and five should remind us that the genome project was only the first step of a very long journey.