You are here: Home: BCU 1|2003: Gabriel N Hortobagyi, MD

	Gabriel N Hortobagyi, MD
Professor of Medicine Chairman, Department of Breast Medical Oncology Director, Breast Cancer Research University of Texas MD Anderson Cancer Center

Edited comments by Dr Hortobagyi

Implications of the ATAC trial in clinical practice

The results of the ATAC trial are quite compelling. Even if you assume for the sake of argument that the curves will come together with further followup, the safety profile of anastrozole is still clearly better than tamoxifen. I cannot prevent endometrial cancer short of removing the uterus, but I can prevent or treat osteoporosis and fractures. Since the safety profile of anastrozole is better than tamoxifen and it is therapeutically superior, I have a problem not offering anastrozole to my postmenopausal patients — not as a neutral choice but as a better choice. I do discuss with my patients the enormous amount of clinical experience we have with tamoxifen, but if my sister developed breast cancer today, I would certainly recommend anastrozole as opposed to tamoxifen.

Use of other aromatase inhibitors in the adjuvant setting

I do not use the other aromatase inhibitors in the adjuvant setting, because there are no adjuvant data. While we have to extrapolate in a number of situations, I do not see an advantage for the other aromatase inhibitors from the existing data. It is possible that some time in the future, someone will show a distinct advantage of one of these other agents, but at this point, the data were generated with anastrozole, so I use anastrozole.

Making clinical decisions in the face of uncertainty

Oncology is one of the classic specialties in which uncertainty is a way of life because of progress and constant change. It is important for our professional organizations to have consensus at various points in the evolution of a particular treatment. Having said that, those guidelines and consensus statements tend to be relatively conservative. Most of the time that is perfectly appropriate, but physicians will have to make individual decisions based on interactions with patients. Some of those decisions will follow guidelines, while others will not.

There are situations where departing from a guideline is clearly wrong. For instance, there is widespread acceptance that aromatase inhibitors should not be used in premenopausal women. Departing from the guidelines in that setting is clearly inappropriate because you would actually reject scientific evidence as the basis for your decisions. But, in situations where there are data and evidence to support various options, there is nothing wrong with deviating from a consensus statement as long as it is appropriate for that specific patient.

Recommending adjuvant anastrozole based on early trial results

There is no comparable trial in the history of medical oncology or breast cancer, and there is no other tumor type with so many well-planned clinical trials conducted. We are in a leadership position in oncology, and we can’t advocate doing the best trials and then ignore the results of those trials. Every single trial we do brings with it some of the unknown.

We started to move over to tamoxifen well before we had five-year follow-up. I remember when Michael Baum presented the early data from the NATO trial in 1982. It had less than two years of follow-up, and he was already publicly talking about the advantages of adjuvant tamoxifen — and the NATO trial pales in size and design in comparison to the ATAC trial.

We have very compelling data about anastrozole from the ATAC trial, in terms of its therapeutic and safety profile superiority. I would be doing a disservice to my patients who are candidates for adjuvant aromatase inhibitor therapy by not presenting the data. I also present tamoxifen as an option, but in the last six months I would say that 60 percent of my postmenopausal patients chose anastrozole rather than tamoxifen. There is no right or wrong decision, but for me, there are compelling data to prefer anastrozole.

Incorporating early research results into practice

I was actually one of the individuals who initially fought against the use of adjuvant tamoxifen — especially in premenopausal women — in the 1980s. Up until the early 1990s, in our own clinical trials at MD Anderson, we did not include endocrine therapy in the adjuvant treatment of premenopausal women. We learn from history that we probably fail to understand the impact of emerging data on the lives of women. Coming from that background, my flexibility in accepting the new and relatively early data of the ATAC trial is more significant to me. If I had understood the deep implications of what tamoxifen could do in terms of saving lives in the early 1980s, I could probably have modified many of our own activities and policies during the subsequent ten years.

There are situations in which one needs to be much more conservative. I was much more cautious in the high-dose chemotherapy area, because there was much to lose. However, the safety issues of high-dose chemotherapy in the 1980s and of aromatase inhibitors in the 21st century are so enormously different that we cannot even compare them.

Providing patients with treatment options and recommendations

One of our major goals is to fully educate our patients by giving them relevant, accurate and complete information, so that they understand their prognosis, treatment options and the benefit-to-risk ratio they will face with each of those options. But we can’t stop there. We also need to make a recommendation after that education. Obviously this recommendation will incorporate our biases and prejudices, but we are better qualified — even with those biases and prejudices — than a patient who just had “oncology 101” during the previous 20-30 minutes. I feel very strongly about that.

Over the past 30 years in medicine, we have moved from a paternalistic approach to the other extreme. Many of my colleagues try to be so neutral that they do not make a recommendation. The burden of decision-making has been removed completely from the physician, who is best qualified to make that choice or recommendation, to the patient, who sometimes is — but most of the times is not — in the best position to make that choice without guidance. I understand and agree that patients need to have autonomy. We clearly have the obligation to inform them fully, but I think we need to go beyond that. We have to get to know our patients and understand their motivations, their understanding of risks and benefits, their definition of therapeutic gain and their acceptable level of risks and side effects. As physicians, we need to help them make a decision. To abrogate that responsibility is an unfortunate — and I hope temporary — trend in the medical profession.

Adjuvant randomized clinical trials of trastuzumab

There is a substantial body of data suggesting that while there is a slight excess in cardiac events with trastuzumab and anthracycline-based chemotherapy in the adjuvant setting, it is unlikely to affect survival in any major way. We have elected to support the BCIRG adjuvant trial, but the NCCTG and NSABP trials are equally worthwhile. All three will contribute to our understanding of how best to incorporate trastuzumab in the adjuvant setting. While it is reasonable to ask whether we can derive the same amount of benefit from trastuzumab with a non-anthracycline-containing regimen as with an anthracycline-containing regimen, the bulk of the data from retrospective analyses of many of our previous trials points to the importance of anthracyclines precisely in HER2-positive patients.

I think it is shortsighted to abandon anthracyclines on the basis of a potential risk for toxicity. The history of oncology is full of toxic agents that were almost discarded until someone found a way to administer them safely. That is true for anthracyclines, cisplatin, alkylating agents and taxanes. We should not be surprised that these drugs have toxicity, but we should not discard them lightly. We should learn to use them safely through clinical trials, and there are many ways to address this issue of developing the safest and most effective combination with trastuzumab.

Dosing and scheduling of chemotherapy

The expression “where there’s smoke, there’s fire” applies to an issue we have been studying for decades — chemotherapy dose and schedule. We learned the hard way with high-dose chemotherapy and bone marrow transplantation, but I think there is room to test various parts of that hypothesis. Several interesting trials are exploring the question of dose-density versus dose-escalation or a bolus of single agents versus combination chemotherapy.

Metronomic chemotherapy was resuscitated in the process of developing angiogenesis inhibitors. Anumber of investigators found that fairly traditional chemotherapeutic agents have an antiangiogenic effect and substantial antitumor activity when given chronically in low doses as opposed to intermittently at high doses.

The experience with taxanes and fluoropyrimidines highlights the importance of scheduling. It is quite likely that the doses and schedules initially approved for both taxanes might not be optimal, and there may be other less accepted or unexplored schedules that might lead us to better administration of these drugs.

Capecitabine: A targeted chemotherapy

Capecitabine is a fascinating agent, which in addition to teaching us more about the fluoropyrimidines in general, brought out the targeted aspect of chemotherapy. Conceptually, capecitabine is a hybrid of a traditional cytotoxic agent and a targeted agent activated on site. This is an absolutely fascinating observation, not dissimilar to the aromatase inhibitors, which also utilize the mechanism of targeting an area rich in the enzyme relevant to the intervention. We have a lot more to explore in this area.

Neoadjuvant trial of capecitabine-docetaxel

As a group, we reached the consensus that for patients whom we are reasonably certain will receive chemotherapy, we prefer to administer all chemotherapy before surgery. In a recent neoadjuvant study, we found that 12 cycles of weekly paclitaxel were better than four cycles of three-weekly paclitaxel followed by four cycles of FAC.

We recently activated a trial of neoadjuvant docetaxel and capecitabine. This trial will compare four cycles of the capecitabine-docetaxel regimen to 12 weekly cycles of paclitaxel, with both arms then receiving four cycles of FEC. We feel that we are building upon the best arm of a previous regimen while also exploring the interaction between capecitabine and docetaxel.

Neoadjuvant chemotherapy in women who may be nodenegative

I am comfortable offering neoadjuvant chemotherapy to women who may be node-negative, because in the past 30 years we have learned that most things in breast cancer are not black and white, but rather gradations or trends. Nothing at this point tells me that node-negative breast cancer is different from node-positive breast cancer with one positive node. If it is appropriate to use a taxane in the adjuvant setting for a 2.5 cm breast cancer with a single positive node containing a 7 mm metastatic deposit, I don’t see a major biological difference for that same primary without that metastatic deposit in a single node.

I probably would not use a taxane off-protocol in a patient with a 1.2 cm node-negative primary, but this too is an evolving area. The initial trials of taxanes were done in node-positive breast cancer, but we are in the process of testing them in node-negative disease. The question will be how to select those patients who should receive everything, those who shouldn’t receive anything, and how to define the grades in between.

For the trials we are conducting now, we do fine needle aspiration of palpable nodes prior to preoperative chemotherapy, so that we know if there is a node containing malignant cells. But, if I have a patient with a welldefined 3 cm breast cancer, I'm going to give chemotherapy whether she has positive nodes or not. Her risk of recurrence is very similar to that of someone with node-positive breast cancer.

Chemotherapy in premenopausal women: Benefits of ovarian ablation

Ovarian ablation with chemotherapy is an area we have not explored adequately. It is certainly apparent that for premenopausal, estrogen receptorpositive patients, ovarian suppression is beneficial. The evolving LHRH analog data suggest that ovarian suppression or ablation need not be permanent. Even temporary suppression has a substantial therapeutic benefit, although we do not know the optimal duration of suppression.

If we accept that this is the case, it is important to develop cytotoxic regimens that do not cause permanent ovarian ablation. Since we backed off six cycles of cyclophosphamide to four cycles of FEC plus a similar duration of a taxane, our preliminary observation is that fewer patients undergo permanent cessation of menses. So, the incorporation of a taxane might have other circumstantial benefits in terms of fertility as well as the major therapeutic goal.

Evolution of breast cancer treatment

Those who don’t know history are condemned to repeat it. For example, it is fascinating to see the evolution of the St Gallen consensus statements over time. At one point, we essentially said that node-negative patients should not receive adjuvant systemic therapy, but for the most recent one, we did not exclude anyone with invasive breast cancer. Our interventions haven’t changed much during that time, but what has evolved is our understanding of the risks and benefits of treatment and what our patients are willing to accept and, in fact, request. All of this is in constant evolution, and what is absolutely true today may not be absolute tomorrow, and what is totally contraindicated today might become standard of care in just a few years.

Select publications

Home · Search