Barbara Brenner’s Reflections on the 26th Annual San Antonio Breast Cancer Symposium—Day 2
(Thursday, December 4)
Abstracts for presentations made at the San Antonio Breast Cancer Symposium can be found at: http://www.sabcs.org/, by clicking on "Abstract Online 2003"
I think it usually happens that, by the second day of this meeting, my mind is spinning with many thoughts, and that’s certainly the case this year. As an introduction to these many thoughts, I should probably give some context of the things I heard today.
Apoptosis: Cell Death and Private Profit
The morning plenary presented by John Reed, President and CEO of the Burnham Institute focused on apoptosis-based therapies for breast cancer. [No abstract provided.] Apoptosis is the process by which cells die. Since the process is interrupted in cancer cells, researchers are looking for ways re-establish apoptosis, thus killing cancer cells. While I might be able to explain the various proteases that Reed and his lab are looking at for apoptosis-based therapy possibilities, the important thing to know about them is that the ones he is most excited about are in pre-clinical phases of development, so nothing will be arriving in your doctors office soon, if at all, from this research. Besides the fact that it’s usually the pre-clinical stuff that is most exciting to researchers, I was struck by the fact that this research is entirely about drug development. While, in and off itself, that’s not a bad thing, it’s being done at the Burnham Institute in Southern California with support from taxpayer funded institutions like the California Breast Cancer Research Program and the NCI. And you can bet that the drugs that are developed—if they are developed—will be in private hands, for private profit. There is something desperately wrong with this picture.
Gene profiles and targeted treatment: we’re not there yet
In contrast to Dr. Reed’s presentation, Soonmyung Paik’s presentation [Abstract 16] on using a multi-gene assay for predicting recurrence in node-negative patients treated with tamoxifen (NASBP Trial B-20) promises to have this test in your doctor’s office any minute. Using an array involving 16 breast cancer prognostic genes, Dr. Paik, Director of Pathology at the NSABP and his colleagues at Genomic Health (a private for-profit company) developed a recurrence score, which they say they validated prospectively (even though all of the patients had long-since been treated). Dr. Paik showed that the recurrence score provides and accurate and precise continuous curve for predicting the likelihood of a distant (non-local) recurrence in node-negative, ER+ patients treated with tamoxifen. This public/private partnership plans to do other studies looking at the benefits of chemotherapy and of tamoxifen using the recurrence score as a predictor. They have not applied their scoring system to the placebo group in the NSABP trial, so they can’t tell the real effect of the tamoxifen in the trial. And, as a questioner noted, in the group that is low risk using the recurrence score from the gene assay, the score misclassifies 10% of patients, meaning that the score is not perfect and it doesn’t tell which patients will recur. In answer to that comment, Dr. Paik acknowledged that we may never be able to tell an individual patient whether she will recur. So much for one of the holy grails of gene research in cancer.
This study was reported in the New York Times under the headline “Test May Aid Chemotherapy Decisions,” and a lead sentence that read: “A new genetic test can help predict whether breast cancer will recur, providing a way to help women decide whether they need chemotherapy.” Genomic Health expects to have the test on the market by early, 2004. Yet the next presentation was from M.D. Anderson, which tried unsuccessfully to validate the recurrence score system looking at untreated patients. In the M.D. Anderson study [Abstract 17], there was no statistically significant correlation between distant recurrence-free survival and the recurrence scores. So much for gene arrays as predictors of recurrence, at least at this meeting.
African American women and breast cancer
One of the most important studies reported here was the outcome of a Southwest Oncology Group (SWOG) analysis of how African American women have done in adjuvant therapy trials. [Abstract 21] Controlling for every variable that has been offered as explaining the different outcome for African American women—including socio-economic status, body mass index, income, and education—the researchers found that both pre-menopausal and post-menopausal African American women still do worse than others treated for breast cancer. As researcher Kathy Albain, of Loyala University, put it, African American women have statistically inferior outcomes. Future SWOG studies will focus on biological explanations for the racial differences in outcome. I asked whether the pre-menopausal African American women did worse than the post-menopausal African American women, but was told that the way the trials were conducted did not permit them to tease out this information.
Looking at the present from the past
The William McGuire Memorial Lecture is usually the highlight lecture of this meeting. This year it was presented by Umberto Veronesi from the European School of Oncology. While his title was “Breast Cancer, A Look to the Future,” it was mostly a trip down memory lane. Since Veronesi has been involved in many significant advances in breast cancer treatment over the last 40 years, he was well equipped to talk about them, and he did. He talked about the shift from Halsted radical mastectomies to lumpectomies and radiation. He said out loud from the podium that radiation following lumpectomy does not change survival as compared to lumpectomy with radiation or mastectomy. He talked about the shift from axillary node dissection to sentinel node biopsy. He talked about using the internal mammary node for staging rather than the axillary nodes. He talked about the shift from extensive field radiation post-surgery to trials he is currently conducting giving a single dose of radiation at the time of surgery. (Called partial breast irradiation, he noted that this is not yet the accepted standard, and should only be done in clinical trials.) Veronesi’s take home message was this: the move to minimum effective treatment will lead to better quality of life, which will motivate women to get earlier detection, which will, in turn, increase survival. The premises underlying that message are pretty amazing when you think about them. Not only does he assume—without providing evidence—that women previously have stayed away from treatment because of quality of life concerns, but he also advances the myth of early detection. (See BCA’s policy on breast cancer screening) So much for the future of breast cancer; what we really got was the future as seen from the past.
Komen Awards: Ignoring environmental links to breast cancer
We also heard the Komen Brinker Award presentations for basic science and clinical research. The most interesting part of these presentations was hearing Susan Braun, the CEO of Komen say that Komen is funding cutting edge research into what causes breast cancer so that the disease can be eradicated. This was a fascinating statement from the same person who is quoted in October Redbook Magazine as saying that nothing in the environment causes cancer. Her view was by implication confirmed by Walter Willett, the clinical research awardee from Harvard, whose focus is on modifiable risk factors for breast cancer who assured us that prevention is not a remote possibility so long as women avoid weight gain, HRT and alcohol, and have access to pharmocogical agents that are estrogen agonists (like tamoxifen). He also noted that we should look at diet and lifestyle in children. Not a peep about non-lifestyle environmental links to breast cancer. None of this is really surprising from an organization—Komen—that takes money from many companies that are contributing to the cancer problem either through their products or production processes, and that for years received free office space thanks to Armand Hammer, the company that brought us Love Canal. Nor is it surprising from someone like Willett at the Harvard School of Public Health, which receives substantial funding from the petrochemical industry.
Non-validated end-points and clinical implications
There’s been a bit of talk here about the need to find surrogate end-points for cancer drug trials—endpoints that will permit us to predict DFS or OS without having to do the long and large trials necessary to establish DFS and OS. But it seems that groping for surrogate endpoints is resulting in clinically meaningless studies. The best example, as I mentioned in my Day 1 notes, is the use decline in expression of Ki67—a proliferation gene—as a surrogate endpoint. A number of the studies presented here rely on this endpoint, directly or indirectly, despite the fact that there is no consistent and standard way of measuring Ki67, and despite the fact that no one has every correlated this gene expression to either DFS or OS. So the studies based on Ki67 have no clinical significance whatsoever. In fact, many of the studies that have been presented here have either not considered clinical impacts of the treatments, or have minimized them. In a presentation I missed, researchers showed that docetaxel gave women with metastatic disease almost 3 more months of life than pacilitaxel. As I was entering the lecture hall as the talk was ending, a breast surgeon from San Francisco who was on her way out whispered in my ear that these patients were trading 3 months of life against extraordinary bone pain. Apparently, this wasn’t discussed in the presentation or the questions afterwards. Someone—everyone—needs to remind the presenters at this meeting that what matters to patients is what these drugs do for them and to them.
The breast cancer puzzle
As I listened to presentation after presentation, each addressing some aspect of the medical problem of breast cancer, I’ve been struck by the lack of any coordinating structure for this meeting. There doesn’t seem to be any rhyme or reason to the diversity of studies being produced, and no one here has stood up before the crowd of nearly 6000 and tried to synthesize or organize the information. We have one researcher saying that breast stem cells are the key to understanding metastasis. We have many, many researchers looking at different patterns of genes, all of which may be irrelevant if the stem cell theory is correct. And we have one or two people telling us that genes alone don’t explain anything—it’s the extracellular matrix, or tissue mircro-enviroment, the context in which the cells interact—that ultimately matters. And then there are the folks in search of novel therapies addressing everything from apoptosis to myoepothelial cells. And, of course, there is the search for prognostic markers, an aspect of the gene research that seems to be largely additive. Every new marker seems to be added to all the other markers, with little consideration of how they correlate to each other. Basically we have a very large puzzle, and what we’re hearing here are descriptions of very small pieces of the puzzle, or chunks of the small pieces. No one is talking about where these pieces fit in the puzzle, or what part of the puzzle picture they represent. And there is no puzzle master. What I wouldn’t give for a puzzle master.