Changing the Face of Molecular Medicine

Changing the Face of Molecular Medicine

Lewis Cantley's discoveries in the laboratory are changing the way we think about and treat cancer.

The 2015 Ross Prize in Molecular Medicine was awarded to Lewis C. Cantley, PhD, who serves as the Margaret and Herman Sokol Professor in Oncology Research and the Meyer Director of the Sandra and Edward Meyer Cancer Center at Weill Cornell Medical College and New York–Presbyterian Hospital. Dr. Cantley received the award at a scientific symposium held at the Academy on June 8, 2015, in his honor. Early in his career, Dr. Cantley discovered phosphatidylinositol-3-kinase (PI-3K), an enzyme that is important for cell growth, insulin signaling, and immune cell function. Dr. Cantley's discovery has led to one of the most promising avenues for the development of personalized medicine. Currently, Dr. Cantley's lab is investigating new treatments for diseases that result from defects in PI-3K and other genes in this important metabolic pathway. Recently, he graciously agreed to sit down with us to discuss his prestigious award and the past, present, and future of cancer treatment.

New York Academy of Sciences: What is the current focus of your laboratory?

Lewis Cantley: My laboratory is trying to understand why cancer cells have altered metabolism and take up significantly more glucose than normal cells. I initially became interested in this area following our discovery of phosphoinositide-3-kinase (PI-3K), an enzyme that is important for cell growth. We came to the realization that when PI-3K is activated, cells consume glucose at significantly higher rates, which is consistent with the Warburg Effect, first described decades earlier by Otto Heinrich Warburg. [The Warburg Effect is the observation that cancer cells produce the majority of their energy by glycolysis and lactic acid fermentation, as opposed to oxidation of pyruvate in mitochondria, as is observed in healthy cells.]

Mutations in PI-3K and other metabolic genes can cause cancer cells to take up increased amounts of glucose, and understanding this process will hopefully reveal new targets for cancer therapies. Together with Craig Thompson and Tak Mak, I co-founded a company called Agios Pharmaceuticals to further explore this concept. Independent of Agios Pharmaceuticals, my lab continues to investigate the mechanisms of altered cancer cell metabolism, and it is our goal to develop cancer drugs for the targets that we discover.

NYAS: Who were your role models in science and how did they inspire you?

LC: Harold Varmus and Michael Bishop were two of my major role models because of their elegant studies on how viruses cause cancer. It was this work that led to the realization that cancer is caused by mutations in human genes. It was paradigm-shifting science because it made us understand that cancer is driven by sporadic mutations in DNA and that the changes in metabolism that Otto Warburg originally observed were a consequence of mutations in genes (like PI-3K) that control metabolism through complex signaling networks.

NYAS: What led to your discovery of PI-3K?

"Understanding the mechanisms by which mutations in PI-3K and other metabolic genes can cause cancer cells to take up more glucose will hopefully reveal new targets for cancer therapies."

LC: The discovery of the Warburg Effect made scientists examine changes in cancer cell metabolism. Much of the 20th century was spent trying to understand how cancers change their metabolism, specifically how they perform anabolic processes at a higher rate. In the late 1970s and early 1980s, work from a number of labs led to the discovery of important oncogenes. In our early work we used viral oncogenes to discover PI-3K. By immunoprecipitating oncoproteins we were able to isolate PI-3K, and at first we believed PI-3K was producing the well-known lipids, PI(4,5)P2 or PI(4)P. However, once we characterized the product, we found out it was chemically distinct from the two well-known phospholipid forms in that the phosphate was on the 3 position of the inositol ring rather than the 4 or 5 position. We were extremely excited since this species had never previously been described.

NYAS: Upon your discovery of PI-3K, did you realize how complex the signaling cascades were?

LC: Our work revealed that PI-3K phosphorylates the 3 position of phosphatidylinositol; however, after that initial discovery we realized that many other phosphorylation combinations could be generated by PI-3K. Sure enough, in subsequent years, a whole new group of lipids was discovered, including PI(3)P, PI(3,4)P2, PI(3,5)P2 and PI(3,4,5)P3, although at the time it was not clear what they were doing. Now we know that many of these lipids are important in cells for controlling protein kinase cascades and actin rearrangement, which is critical for cell movement.

I was extremely excited by the importance of PI-3K for human disease. Initially our team was mainly focused on insulin signaling rather than on cancer, but soon we realized that there were commonalities between insulin signaling and the evolution of cancers. The story of PI-3K has certainly turned into a bigger story than I could have ever anticipated.

NYAS: PI-3K inhibitors work quite well in blood cancers, but show more variable results in solid tumors. Why do you think that is?

"We were extremely excited when we characterized the position of phosphorylation and realized that this species had never previously been described."

LC: The PI-3K gene that is mutated in solid tumors (PIK3CA) encodes the same enzyme that insulin activates so inhibitors of this enzyme cause insulin resistance resulting in hyperglycemia, which limits the dose of drug that can be used for therapy. In contrast the PI-3K inhibitor that was approved for treating B cell lymphomas, idelalisib, targets the enzyme encoded by PIK3CD, which does not mediate insulin responses. Thus there is less toxicity and higher doses of drug can be achieved, allowing more effective killing of tumor cells. I also think that the total number of cancer cells in the body at the time a patient goes on therapy has a major role in explaining resistance to therapy. We now know that there is tremendous heterogeneity in the mutational events in most solid tumors and the more cells present, the more likely that a few cells in the tumor will be resistant to the therapy. That is why we are exploring the usefulness of neo-adjuvant therapy, the delivery of an anticancer drug prior to surgery. Another option for improving patient outcome is adjuvant therapy, the delivery of an anticancer drug immediately following surgery, even before recurrence is detected.

Often by the time metastatic cancer is diagnosed, the total number of cells in the body can be massive. Bert Vogelstein aptly pointed out that every time a cell divides there is a chance for an error in DNA replication, resulting in genetic aberrations, and the more times that happens the greater the diversity of mutations in the tumor and the lower the probability that a single agent will kill all cells in the tumor. Initial clinical trials in solid tumors are typically done in patients who have metastatic disease and have failed multiple therapies, so it is a high bar to achieve complete responses in this setting.

NYAS: Why do certain cancer drugs look quite promising in preclinical models yet do not perform as well in humans?

LC: New cancer drugs are often tested in mice that have a single, small tumor. Since the tumors in mice contain relatively few cells, the odds that we can kill all those cells are rather good. The clinical setting with human patients is far more challenging and complex because, as I indicated before, human cancer cells have greater genetic diversity and there are at least 100 times more cells than in a mouse tumor. That is not to say that mouse models are bad, but we need to pay better attention to the mathematics. In normal preclinical studies we give seven mice the experimental drug and seven mice receive the placebo. As pointed out by Bert Vogelstein, these numbers are far too low. We need to increase the number of animals used in preclinical studies and focus on therapies that cure all the mice, then we are far more likely to find drugs or drug combinations that are also effective in humans.

NYAS: If you had a crystal ball that showed you the future of cancer research and treatment, what would you like to know right now?

"We know that we can pick up circulating mutant DNA in the case of metastatic disease, but it would be fantastic to do this for very early stages of cancer."

LC: That's a tough question! One of the things I would like to know is whether we will have technologies available in the future to detect circulating mutant DNA at very early stages of disease. I think it would be great to have a test that would allow us to intervene with therapies potentially even before a tumor can be felt by a patient or detected by standard imaging techniques. A test like this would have to be extremely sensitive so that we could detect extremely low levels of circulating mutant DNA. We know that we can pick up circulating mutant DNA in the case of metastatic disease, but it would be fantastic to do this for very early stages of cancer.

NYAS: Your clinical test sounds like a fantastic idea—what are the pros and cons?

LC: If we were able to develop a test like this and it were cost-effective, it could very well become a routine clinical procedure that takes place during the annual physical every year after the age of 50. If people are at high risk for cancer, they could have the test done starting at age 30. These test results could potentially tell you that you have circulating copies of oncogenic mutant DNA. I believe that if clinicians administered targeted cancer therapy at these early stages of disease, we would have a much higher likelihood of a cure. The success of this whole plan depends on the development of targeted cancer drugs that are safe and have few off-target effects. Developing these drugs and testing their safety could take as long as 5–10 years. Most of the drugs we currently use for cancer therapy would not be acceptable to use in this setting since they could cause more harm than good and even cause new cancers to occur. Another caveat to this blood test is the possibility of false positive results, where patients may show the mutant DNA but never actually progress to full-blown disease. I think that personalized medicine is the future. If we truly want to cure cancer, we need to target the cancer cells more effectively and hit them earlier with safe, non-toxic drugs.

NYAS: PI-3K is at the interface of insulin signaling and cancer; what is the relationship between these two?

LC: Many types of cancer cells express higher levels of insulin receptor (IR) or insulin-like growth factor 1 receptor (IGF1R) than the tissue from which they evolved. If a patient with this type of cancer becomes insulin-resistant, as could happen from a high-sugar, high-carbohydrate diet, there will be high levels of circulating insulin and IGF1in the blood. This is a very dangerous situation because if the tumor expresses IR or IGF1R, it will be getting a strong signal for activating PI-3K all the time, even if PI-3K is not mutated. This will drive tumor growth and may render the tumor less vulnerable to chemotherapy. If I had a cancer that expressed high levels of IR or IGF1R I would go on a low-carbohydrate diet the very next day.

"If I had a cancer that expressed high levels of IR or IGF1R I would go on a low-carbohydrate diet the very next day."

High levels of dietary sugar can cause insulin-resistance, which results in near-constant elevation of circulating insulin. We know that insulin activates PI-3K, which is almost certainly driving a large fraction of cancer growth. In the United States there is a very high fraction of people who are insulin-resistant, but many of them are undiagnosed. It is a frightening possibility that we will retrospectively regret making sugar cheap and broadly added to foods the same way we now regret making cigarettes cheap and broadly available 70 years ago.

NYAS: What does winning the Ross Prize in Molecular Medicine mean to you?

LC: I am tremendously honored and excited to win the Ross Prize. I am particularly grateful for this award because it is not given for a single discovery, but rather a body of work where a discovery has been translated into a clinical outcome. That is difficult to do; but I certainly did not do that alone. Hundreds of people collaborated with me at various stages—from the mouse models, to the biochemistry, all the way to carrying out a clinical trial. I have been very fortunate in my career to work closely with passionate people who are focused on a common goal of identifying new cellular targets for cancer drugs.

About the Ross Prize in Molecular Medicine

The annual Ross Prize in Molecular Medicine was established in conjunction with the Feinstein Institute for Medical Research and Molecular Medicine. The winner is an active investigator who has produced innovative, paradigm-shifting research that is worthy of significant and broad attention in the field of molecular medicine. This individual is expected to continue to garner recognition in future years, and their current accomplishments reflect a rapidly rising career trajectory of discovery and invention. The winner receives an honorarium of $50,000.

Previous Award winners include: John O'Shea, MD, National Institute of Arthritis and Musculoskeletal and Skin Diseases (NIAMS), National Institutes of Health (2014); and Dan Littman, MD, PhD, New York University (2013).

An open-access eBriefing is available for the 2015 Ross Prize in Molecular Medicine award symposium and can be viewed at

For more information, please send an email to