The End Of The Traditional Cancer Clinical Trial: N-Of-1 Is The Future
By
Leonard Zwelling
https://ascopubs.org/doi/10.1200/JCO-25-01453
I may be an oncologist officially because I took and passed a board exam in 1983, but the truth is that I am more a student of clinical oncology from afar and always have been. I loved internal medicine, but emerging from the Duke Medicine residency program, I had to choose a sub-specialty like all my fellow residents and, for many bad reasons, I chose oncology. I just don’t have the personality of a clinical oncologist which is what my wife told me on our first date. On one hand it was good I didn’t listen given my subsequent career. On the other, that doesn’t make her wrong. She rarely is.
In the 1970s, it was a good time to pursue sub-specialty training at the NIH. Clinical work there qualified you for internal medicine and sub-specialty boards in only three total post-graduate training years. This was called short-tracking. In addition, after a year of clinical work, two additional years of laboratory time were guaranteed thus preparing us NIH fellows for careers in academic medicine. There was also bonus pay in the third year to keep us in Bethesda. It worked for me because I stayed for nine years, published forty of so papers, attended on the clinical service of the Medicine Branch of the National Cancer Institute once a year, and eventually became a Senior Investigator, part of the tenured staff. This all made a great launching pad for my subsequent career at MD Anderson.
Also, as an NIH fellow, I was in the U.S. Public Health Service, an official uniformed service. Thus, this fulfilled my national commitment that precluded my eligibility for the draft and service in Vietnam. So, following two years of internal medicine training at Duke, I trained like an oncologist, but really became a biochemist and molecular pharmacologist. No one would have confused me for a real oncologist when I got to MD Anderson in 1984. And, by the way, no one ever did.
Much later, after business school, I oversaw the infrastructure for research at Anderson, with a focus on research that was federally-regulated. This included the process by which clinical protocols were reviewed, approved, and funded. Overseeing the clinical research infrastructure gave me additional insights into the methods of clinical cancer research from 1995-2004. It also got me interacting closely with the greatest clinical cancer investigators in the world at the time, the faculty of the Division of Cancer Medicine. Every single department chair was a world’s expert in his or her chosen area. For most of those years, Dr. Hong was the Division Head and he was a world-class investigator himself.
Since then, I have been watching oncology from the sidelines. I still read The New England Journal of Medicine (in its printed form) and try to keep up with the newest developments in cancer research. I really can’t, but I do my best.
I am stunned by the effectiveness of new oncogene-product-targeted drugs as well as the power of the new immunotherapy. There is rarely an issue of the NEJM when a new, effective pharmaceutical is not identified. Dr. Kleinerman always advocated for immunotherapy. She was just forty years ahead of her time.
One of the disciplines that I studied as a vice president was the design of clinical trials and the circuitous route faculty had to take to get novel treatments to patients, especially if the trials being proposed varied in any way from the commonly employed designs of the time. It was all about phase one, phase two, and phase three designs in groups of patients that were considered to have the same cancer type.
Today, we know that this design is completely wrong. Patients are not the same and far more importantly, neither are their cancers even though those cancers may look similar under a microscope and present with a seemingly common, well-recognized set of signs, symptoms, and traditional lab test results, like findings on a CT scan.
Thus, the traditional model used to assess the safety and efficacy of novel anti-cancer agents is of little use because patients’ cancers cannot be grouped as if they were alike. At the level of the genome, they are not alike. In a group of twenty non-small cell lung cancer patients, each could be malignancies with different genetic profiles. How can they be treated alike?
Instead, as the attached article from a good friend and loyal blog reader, Razelle Kurzrock, demonstrates, the way to treat cancer in 2026 is to assess as many possible targets for novel gene-product-directed agents in any given cancer and devise a treatment strategy addressing as many as possible. In other words, lung cancer is simply not a diagnosis. Tell Dr. Kurzrock what genes are being aberrantly expressed in a patient’s cancer and she can tell you how to get the best result for that patient. All clinical trials will have an N of one. As Dr. Kurzrock wrote me, it is the algorithm used to select the treatment that is tested in these new N-of-1 studies.
In my wildest dreams as a Duke intern in 1973, I could never have imagined such a thing. I couldn’t even have thought it possible twenty years ago as a vice president. Yet the reality unearthed by modern cancer biology, biochemistry, and genetics dictates a new approach to the design of cancer clinical trials.
N-of-1 trials are not just a new way to explore novel therapies. They are the only way that makes scientific sense. Now all we need is for all the editors of all the journals to realize these trials are at the cutting-edge of clinical research and for the FDA to devise a way to both approve the trials and also use the resultant data to gain market approval for new drugs discovered to be beneficial in this way.
You personalize your iPhone. Why not your cancer therapy?
