In the last 20-30 years, our understanding of chemistry that controls life has expanded beyond the range that even most scientists anticipated twenty years ago. Consider that the focus of much discussion in bioscience in the early 1990s was whether enormous effort should be put into sequencing "the human genome." After the then anticipated 20 year project was finally initiated, it was actually completed about 7 years ahead of schedule. Today, just 10 years after sequencing the first human genome—about 3 billion bases of DNA—we can sequence any individual's (or animals) genome in just a few days. In fact, sequences from over a 1000 human genomes were released last year as part of a project to analyze the variation of human DNA around the world.
However, DNA sequencing is just one example. Routine laboratory methods that were just being thought of twenty years ago to analyze many thousands of genes, proteins, metabolic intermediates, and other biological molecules in living systems are in common use. Labs do in a week, or sometimes a day, things that I and my co-workers spent years of graduate work doing.
An Avalanche of New Data
Biology, particularly, molecular biology, has transitioned into a whole new era over the past quarter century to produce databases overflowing with data about which genes are on or off, which proteins interact with each other, how do levels of RNA change, and which enzymes are active in thousands of biological systems under all different conditions. The problem now is trying to understand what all these results mean and untangle which one are the important events that cause or control a response, and produce the rest of the changes.
Meanwhile, as researchers continue to struggle with massive amounts of data they mostly don't understand, results work their way to your doctors, hospitals, and clinics in a trickling drop-by-drop manner. An accompanying article discusses some challenges have arisen as DNA sequencing has finally begun it long anticipated move from a research technique to a medical diagnostic. However, the real challenge with bringing the advances of molecular medicine is the lack of a bridge to link the dynamic advances being made in molecular research with practical clinical treatment.
Sorting Out What New Discoveries Mean to Patients
A doctor isn't simply interested in what DNA mutations or protein markers a patient's tumor has, but rather, what the mutations indicate about which drugs the tumor would be most vulnerable to. However, finding out this information requires wading into a vast and changing sea of primary research. Many thousands of labs are making new discoveries every day. To make use of these discoveries, healthcare providers and testing laboratories must have access to the current findings, and herein lies the problem.
As discussed in a related article, traditional peer-reviewed scientific publishing is simply not an effective way to transfer the latest research to the clinical use. Also, as treatments get more complex and personalized, researchers get little visibility as to what the patients' responses are to drugs and treatments. This detailed information about patients' experiences with different therapies is vital to setting up new studies that will help understand surprising or resistant responses and improve therapeutic targeting overall.
The Closer We Look at Cancer, the More Cancers We See
For example, molecular analysis is making huge advances in cancer diagnosis and treatment. It is changing the way tumors are categorized and the understanding of how the disease progresses. One recent study from the Cancer Genome Atlas Network looked at the molecular features of tumors from over 500 patients, and identified four major classes of breast cancer based on the activity of certain genes. However, those general groups all broke down into multiple sub-types because tumors in each of these four classes had various combinations of mutations in 23 different genes. In other words, different combinations of changes produced similar tumors. Drugs that work against cancer cells with genes altered one way, may not work against those altered another way.
This sort of molecular analysis is now going on with all types of cancers, as well as some other diseases. It has really only been possible in the last few years and it is the type of data that's required to really advance understanding of how these diseases work. The problem, though, is that, as more and more distinctions are identified, it becomes increasingly difficult to understand what different combinations of features mean in terms of practical courses of action for treatment. What are the best treatments for the hundreds of permutations of the 23 different possible mutations identified in the study above? Which should be used to decide the treatment for a particular cancer? That's the scope of the problem, and it is just the first problem. The second issue is how will research and medicine will manage development, approval, and administration of the multitude of treatment variations required to target hundreds of tumor variations.
Changes Needed to Keep Everyone Up to Speed on It All
In an increasingly complex taxonomy of disease where new findings crop up almost daily, physicians and patients need to know what data needs to be gathered about their specific disease variation and what the results mean. However, as the quantity of research findings grow about different cancer variations, it is difficult for even cancer specialists to stay up-to-date with the latest findings in their field, much less actually integrate these results into practical treatment scenarios. Yet, for patients, some new findings could literally provide life-saving information.
For cancer research and treatment, one group, Cancer Commons, a non-profit organization based in Palo Alto, is taking on this problem head-on. They are trying to create "rapid learning communities" that link patients, physicians and scientists, and remove the divisions between cancer research, cancer treatment, and patients' experiences. Their goal is to build an infrastructure to integrate data from all three sources into a single accessible resource which can provide the information for more precise effective treatments while also giving researchers more direct access to clinical results and patient feedback to help drive the direction of their future research.
Cancer Commons is an ambitious effort but it envisions the sort of change required. Bridging the expanding gap between new findings and clinical practice is a challenge that must be overcome to make "personalized medicine" a reality. For this happen, requires not only changing the way medicine is done, but also changing the way science is done.