Does Breast Cancer Risk REACT to Celecoxib?

Charles Coombes, MD; Kathy D. Miller, MD


December 27, 2017

Kathy D. Miller, MD: Hi. I am Kathy Miller, professor of medicine at the Indiana University School of Medicine in Indianapolis. Welcome to Medscape Oncology Insights, coming to you on a very rainy day from the 2017 San Antonio Breast Cancer Symposium (SABCS).

We have wondered for a long time about inflammation and its association with cancer. Some of the early textbooks called cancer the "wound that never healed," but does inflammation really stimulate the growth of cancer cells? The REACT trial[1] being presented at this meeting was designed to help answer that question. With me today is the primary investigator of that study, Dr Charles Coombes, professor of surgery and cancer at the Imperial College of London and professor of medical oncology at Charing Cross Hospital in London, all in the United Kingdom. Welcome, Charles.

Charles Coombes, MD: Thank you.

Association Between Cancer and Inflammation

Dr Miller: Tell us about the association between cancer and inflammation. It seems almost like a chicken-and-egg question to me.

Dr Coombes: There is increasing evidence that inflammation can get cancer started. We often do not associate inflammation with breast disease because women frequently get cyclical breast pain. And breast cancer often happens to postmenopausal women, but pain is not a feature of precancer or cancer. But in the breast there is increasing evidence that an inflammatory process around the breast tissue can start the cancer process.

Dr Miller: We have seen some tantalizing suggestions from some large population studies, like the Nurses' Health Study,[2] that women who regularly took an anti-inflammatory, aspirin per se, were less likely to develop breast cancer and may be less likely to have a recurrence if they did develop the disease. That has been difficult to study in the laboratory and difficult to prove. How did you tackle that in the REACT trial?

Dr Coombes: We decided to look at celecoxib specially because it inhibits the cyclooxygenase (COX)-2 enzyme. Aspirin is a COX-1 inhibitor, and although that enzyme causes part of the inflammatory process, it does not cause the part that we think is the beginning of cancer. In fact, there are many preclinical models in other forms of cancer, showing that COX-2 inhibition can prevent cancer progression and metastasis.[3,4]

Dr Miller: The data I am most familiar with for prevention with celecoxib are from colon studies, not in breast cancer.

Dr Coombes: Correct. It is one of the indications now for familial polyposis coli after several studies[5,6] have confirmed that inhibition of COX-2 can prevent cancer from starting.

Dr Miller: My memory of those studies was fewer polyps and fewer that progressed to cancer.

Dr Coombes: That is right. Exactly.

Design of the REACT Study

Dr Miller: Tell us about the design of REACT.

Dr Coombes: We randomized women who had a significant chance of recurrence [of breast cancer] to receive a moderate dose of celecoxib (400 mg daily) in a 2:1 design. For every two women who got the active substance, one got placebo. We gave them treatment for 2 years and then followed up. Women who had a very good prognosis were excluded.

Dr Miller: How many women?

Dr Coombes: 2500 approximately.

Dr Miller: That is not a small endeavor.

Dr Coombes: No, absolutely. Initially we wanted a big number because we did not think the effect was going to be too high. Other trials in breast cancer with tamoxifen or Herceptin® (trastuzumab) have been targeted at specific subgroups of women with the [related] receptor, but for the COX antagonist, we really did not have any specific biological marker that could predict the efficacy of the compound.

Dr Miller: Were women eligible irrespective of disease phenotype or did you exclude certain subtypes?

Dr Coombes: We included estrogen receptor (ER)-positive and ER-negative [subtypes], but because of the cardiac issues [associated with celecoxib], we excluded HER2-positive patients. As you know, Herceptin, which is given to that subgroup, is also cardiotoxic in certain patients, so we felt that we could not give the drug along with Herceptin.

Dr Miller: Now we should speak a little bit more about the cardiac toxicity issue because that resulted in the abandonment of another large COX-2-inhibitor trial [with celecoxib].[7] My memory is that that trial was abandoned around the time that a different COX-2 inhibitor ran into cardiac toxicity signals, and with the concern that this might be a class effect, the celecoxib trial was canceled as well.

Dr Coombes: Yes. That was using a compound called Vioxx [rofecoxib], which was used previously for colorectal cancer prevention.[8] It was that compound that made everyone feel that this was a class effect, that celecoxib had to be similar. In fact, subsequent big studies[9] with more than 200,000 patients have shown that although you do get a slight increase in cardiac events, risk with celecoxib is smaller than was previously anticipated.

In our trial, we saw no increase at all in any cardiac or stroke events, which was our main concern.

REACT Results

Dr Miller: Tell me about the breast cancer events. Did you see a signal of activity here?

Dr Coombes: This was interesting. Unfortunately, overall, there is no effect. If you take the entire population, there is no benefit in either delaying time to progression or overall survival. But when we did the subgroup analysis, we found a signal in women who did not have chemotherapy. The subgroup was only 660 patients.

These were women who, on the whole, had ER-positive disease and a good prognosis. Nonetheless, it appeared to be a possibility that in that subgroup, the compound could do some good.

Dr Miller: It is fascinating to get deep in the weeds of subgroups and to wonder what the biology of that might have been. Could it be that the chemotherapy abrogated the inflammation, that there was not an ongoing inflammation left?

Dr Coombes: Absolutely. Our current theory is that chemotherapy has an anti-inflammatory [effect]. For example, some patients with rheumatoid arthritis get chemotherapy.

Dr Miller: Other autoimmune diseases, like lupus and such, may have a holiday.

Dr Coombes: Exactly. We do know that chemotherapy damps down inflammation; and if celecoxib is doing the same thing, then you really would not necessarily expect it to have an additional effect. But it is more complex than that. Over the past 15 years since we started this study, a lot more work has been done to understand the effect of COX-2, and as you know, COX-2 makes a class of compounds called prostaglandins, which are now known to have effects on all cell types.

All cell types produce prostaglandins and most cell types respond to prostaglandins in different ways. We think what is going on is a combined effect, but the drug is preventing prostaglandins from having an action on stromal cells, inflammatory cells, and also breast cancer cells themselves.

Association Between Obesity and Inflammation

Dr Miller: I am going to ask you about another potential association or subgroup. Did you look at the impact of obesity or body mass index? There has been some thinking that perhaps an increase in risk with obesity is also related to higher levels of inflammation.

Dr Coombes: We have not specifically looked at that. It is one of the things that we really want to do going forward, but I cannot tell you that at the moment. What I can tell you is that we do have another set of indicators that will help us choose the patients that will benefit from this drug now, and because we have collected all of the tissue samples from 90% of the patients that entered the trial, we will be able now to go back to analyze them for these markers that have been discovered in the past 15 years or so, to see which group of patients really do benefit.

Suggestions for Investigators

Dr Miller: It will be fascinating work. We will get you back to tell us about that. I have to ask you one other practical question for the researchers listening. I am aware of two ongoing adjuvant trials looking at aspirin—one in the United Kingdom, the ADD-ASPIRIN trial,[10] which is not limited to breast cancer, and a breast cancer–specific trial in the United States. If you were advising those investigators, would anything in your data have you amending the trial, changing eligibility, or changing how those trials would go forward?

Dr Coombes: Until we have sorted out and investigated the relevance of the COX-2 and also the COX-1 signatures, which we can now do using nanostring technology on paraffin sections, I cannot really answer that. It does seem to me to be logical, that since we did not detect any gastrointestinal (GI) effects of celecoxib, it might seem a good idea to combine both agents in women who do not have past history of GI toxicity.

Dr Miller: We are learning that the area of inflammation is clearly important. It is much more important and much more complicated than what we had thought before. Thank you for your work and thank you for joining us today. To you, our audience, thank you for joining us as well. This is Dr Kathy Miller, reporting from SABCS 2017.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.