In this interview we discuss important developments in the diagnosis and management of breast cancer and more.
Oncology (Williston Park). 31(5):346-348.
I. Craig Henderson, MD
1.Dr. Henderson, you have served on the Editorial Board of ONCOLOGY since the journal’s inception 30 years ago. Your distinguished career as a breast cancer oncologist spans more than 4 decades, and includes nearly 300 published articles and numerous clinical trials. What were some highlights of your career as a young oncologist, when you first conducted clinical and translational research at the National Cancer Institute? What discoveries inspired your early work?
DR. HENDERSON: I started my career at the National Institutes of Health, working in viral oncology in Dr. George Todaro’s lab. This is where I got interested in genomics.
My interest in clinical trials was first aroused when I went to Dana-Farber Cancer Institute, particularly working under Dr. Tom Frei. My first two trials involved: 1) testing a (then) new adjuvant chemotherapy regimen Adriamycin (doxorubicin) plus cyclophosphamide (AC for 5 cycles) for early breast cancer; and 2) a dose study with high-dose chemotherapy using a regimen called “super-CMF (cyclophosphamide, high-dose methotrexate with leucovorin rescue, and fluorouracil).” Dose studies and adjuvant therapy have been ongoing themes for the rest of my career. People forget now that using doxorubicin for potentially “curable” patients was once extremely controversial, not just because of heart disease but because of all the other side effects. It was considered too toxic to be used in the adjuvant setting, but it is standard now. Super-CMF was one of the early high-dose trials, developed with the inspiration of Dr. Frei, as well as others. Eventually we tested other doses, both in high-dose regimens using autologous bone marrow transplant and in protocol 9344, which I chaired for the Cancer and Leukemia Group B. Trial 9344 was a study of high-dose doxorubicin, as well as the addition of a taxane. That trial established the role of taxanes in the adjuvant treatment of early breast cancer.
It was also in my early years at Dana-Farber that I began to be aware of the importance of randomization. This was partly due to inspiration from Dr. Marvin Zelen, who became head of the department of biostatistics shortly after I arrived there. That was the basis of my enthusiasm not only for large randomized trials, but also for eventually chairing the Oxford Overview group with Sir Richard Peto.
Another exciting thing in my career was doing the first phase II trial of Herceptin. Dr. Dennis Slamon had done a phase I trial at the University of California, Los Angeles and had gotten a few responses, but the trials we did at the University of California, San Francisco and Memorial Sloan Kettering were the first in which Herceptin was given for prolonged durations. We got magnificent results. Even though the total response rate was small, the patients who did respond were unexpected responders-that is, they had failed everything else, were far along in their disease, and their responses were really quite dramatic and complete. To take a drug that was never used before, and one that is relatively benign and nontoxic, and get these responses was by itself very exciting. When you consider how Herceptin changed the whole field, not just for cancer, but for medicine-the use of monoclonal antibodies, the idea of targeting growth factors-it is clear that this was a transformative moment in medical history.
2. What do you view as the most important developments in breast cancer diagnosis?
DR. HENDERSON: The first development is the stereotactic biopsy. When I started my career, I was a medical student at Columbia University in New York. The most famous and powerful breast cancer surgeon was there at that time, Dr. Cushman Haagensen. This was an era when patients who had a lump were taken to surgery and woke up with the treatment-that is, mastectomy, and usually a Halsted radical mastectomy-done before they even knew the diagnosis. That has changed, and now with stereotactic biopsy we don’t have to admit the patient to the hospital. We can do the biopsy under local anesthesia and the patient then has weeks to make a decision about the right therapy for her.
More recently, the sentinel lymph node biopsy was introduced. We came to understand that removing lymph nodes had, at most, a marginal benefit in improving patient survival. We continued to do lymph node dissections mainly because we wanted information on the potential spread of disease in order to make decisions regarding adjuvant systemic therapy. With the sentinel lymph node biopsy, we could avoid the lymph node dissection, the lymphedema, and other complications.
People also forget that when I started my career there was no estrogen receptor measurement. In fact, we didn’t have a high opinion of the use of hormone therapy. That changed as we began to recognize not only that the majority of patients with breast cancer were estrogen receptor–positive, but also that these patients had a different disease course, a different responsiveness to therapy.
Stereotactic biopsy, sentinel lymph node biopsy, and measurement of estrogen receptor expression represent three revolutionary changes in the diagnosis of breast cancer that have important implications for the treatment and management of the disease.
3. What do you feel are the top three developments in the management of breast cancer with chemotherapy and/or targeted therapy?
DR. HENDERSON: I find it difficult to limit it to three. I look back over my career and cannot believe how substantial the changes are. When I was in medical school, we didn’t talk about mutations as a cause of cancer.
First is the use of adjuvant systemic therapy. The original idea of this was taken from pediatric patients in whom adjuvant therapy was used first, but it was transferred to breast cancer by people like Dr. Tom Frei. It has had as great an impact on survival as any single thing that we have done.
Second is the introduction of tamoxifen. I mentioned that the routine measurement of the estrogen receptor and later the progesterone receptor in the diagnosis of breast cancer were important changes, but the introduction of tamoxifen was also important. Tamoxifen is remarkably well tolerated, and as a result we were able to treat patients for 2, then 5, and then 10 years without many problems. We were able to conduct large trials with prolonged treatment that eventually demonstrated the beneficial effect of endocrine therapy on the survival of these patients. This represented the earliest form of targeted therapy; we were targeting the estrogen receptor.
Third, in the 1990s we started using Herceptin. I think there is no single event that was more important than that-it changed not only the treatment of breast cancer, but all of medicine.
Finally, the thing that I find most exciting and promising now is the introduction of genomic assays. We have just seen the very tip of what this is going to do in allowing us to personalize medicine, so that we have the right treatment for the right patient.
4. As a past chair of the US Food and Drug Administration’s Oncology Drug Advisory Committee, and given your extensive background in breast cancer research, what are two clinical trials that you would like to see conducted in the near future?
DR. HENDERSON:The most important trials are those that further the frontier, that is, trials that introduce new concepts or change our fundamental understanding of the disease. However, there are a couple of more limited, specific breast cancer trials I’d like to see soon.
The first is determining the predictive value of genomic assays for groups other than just node-negative patients. This was a good group of patients in which to initially evaluate genomic assays because the vast majority of node-negative patients were already “cured” and couldn’t benefit from adjuvant chemotherapy simply because their prognosis was too good. But there is a substantial subgroup of node-negative patients who need adjuvant chemotherapy. It was reasonable to start out by focusing on these patients, but I believe we already have sufficient preliminary data that suggest there are other groups of patients for whom these assays will be important. In the long run, once we know which assays to use and how to interpret results from different assays, we will be able to utilize this technology more broadly. We need more studies of genomic assays in node-positive patients, in patients with large tumors, and in other subgroups, because the estimates of the likelihood of recurrence over a period of 5 or 10 years are going to differ in these various subgroups.
Second is better use of neoadjuvant therapy in drug development assays, such as I-SPY. We are moving in that direction, but I believe further randomized trials are necessary to refine the optimal use of I-SPY and other innovative trial designs that will allow us to test concepts in vitro and in the patient-and then use the results of these trials, which have relatively fewer patients and shorter follow-up, to better design our definitive randomized trials.
5.You have served on the Board of Directors of the San Francisco Opera for 17 years, and you sponsored a young artist in residence through their Adler Fellowship Program. In what ways does the opera, or music in general, inspire you as a physician?
DR. HENDERSON: One of the things that differentiates a physician from other scientists is the human dimension. It’s not that other scientists are not interested in human biology or the human implications of their work, but rather that this dimension is the core of what a physician does.
I believe that the arts are very important in expressing those unique qualities that define us as human and help us maintain our perspective. For many years I taught ‘Intro to Medicine’ at Brigham and Women’s Hospital in Boston. One of my sessions was on truth telling, but my interest was more on exploring with the students how a physician knows the truth. I believe there needs to be a tension or war between the different parts of a physician’s brain. One is the scientific part, which processes data and should be objective and extremely rigorous, and the other half is the part that understands what it is like to have these diseases, suffer them, and understand the feelings and reactions of the human being. That is the “fuzzy” part, but that is the part that is augmented by the arts, and, I find, augmented by music in general and, for me, opera in particular.
Physicians are scientists with an intense interest in all things human, including emotions, their patients’ subjective responses to treatment, and the soft and fuzzy elements that are so important in distinguishing us from the rest of living creatures. The arts contribute importantly to these elements of the physician persona.
Financial Disclosure: The author has no significant financial interest or other relationship with the manufacturers of any products or providers of any service mentioned in this article.