Arimidex: Is Longer Better?

Patient Expert

Hundreds of thousands of breast cancer survivors take an aromatase inhibitor (AI) - Arimidex, Femara, or Aromasin - as part of a long-term effort to reduce the risk their cancer will come back. But these drugs are still so new, long-term data that would determine the optimal duration of treatment just isn't available. Should you take your AI for 5 years? Seven? Ten? No one knows. How do you determine your best course of action?

Back in 2005, when I began taking Arimidex (after 3 ½ years on tamoxifen), my oncologist told me I'd be on it for 5 years - after which "we'd see."

"What does that mean?" I asked. "What are we going to 'see'?"

He smiled a bit abashedly and said, "We'll see where the research is." And went on to explain to me that the first large-group studies of survivors taking AIs were, at that point, only about 5 years out.

Did Arimidex appear to be effective? Yes.

Would its effectiveness against cancer recurrence continue once the patient stopped taking it? Don't know. If so, what would the optimal length of treatment be? Don't know.

Would there be long-term negative side effects based on a 5-year course? Or 7 years? How about if you take it for 10 years? Would it help, hurt, do nothing"?

Don't know.

And for all intents and purposes, that's where we still are today.

What's the best length of time to take Arimidex or another AI, balancing effectiveness against recurrence, current side effects, and long-term side effects?

Don't know.

So, we're taking these powerful drugs, and we don't quite know what effect they're having on our health - both positive, and negative?

That's right. And you know what? That's OK.

It takes years to determine just how well any particular drug works. Pharmaceutical companies can spend decades to develop, test, and bring a drug to market- which helps explain why some of these life-saving drugs can be so tremendously costly.

In the case of cancer drugs, researchers are obviously just as eager as Big Pharma to get drugs on the market. A new drug can be the difference between life and death; the longer the approval process, the more people die waiting.

But clinical trials can only be hurried along so much. If one of the goals of the trial is to see whether MiracleDrugA is better than CurrentDrugB after 5 years, well, then you need participants to take the drugs for 5 years, then watch and record how they do - 5, 10, 15, 20 years later.

Luckily, there's some common sense and flexibility in the process. Trials are sometimes revised midstream, when one arm shows results too positive to ignore - especially in cancer trials.

That was true of the large ATAC trial, begun in 1996 and completed in 2010. Designed to determine which protocol worked better - Arimidex, tamoxifen, or a combination - it became apparent 6 or 7 years into the study that women taking Arimidex (whether alone, or with tamoxifen) were seeing fewer cancer recurrences than women taking just tamoxifen. At that point, all the women in the trial began taking Arimidex - and the trial shifted focus to determining whether Arimidex alone, or in combination with tamoxifen, produced better results.

So, when will we know the optimal duration of treatment for our AIs?

A clinical trial testing different durations of Arimidex treatment, run by the National Institutes of Health and sponsored by AstraZeneca (the manufacturer of Arimidex) is forecast to be completed in 2015, with results probably available in 2016, and recommendations to follow. However, the test is 3 years vs. 6 years; not much help for those of us asking ourselves, "Is 7 years better? How about 10?"

If results show that 6 years of Arimidex are more effective than 3 years, we might surmise that 7, 8, or 10 years would be better still - though the data won't be there to prove or disprove that assumption.

So, given the dearth of data, why are some oncologists advising their patients to continue taking Arimidex or other AIs indefinitely?

Because medicine is as much art as science.

Two years ago, five years after I'd started taking Arimidex, I reminded my oncologist that we'd arrived at "wait and see" time. Should I continue taking it? Would continuing lessen my chance of recurrence? Would it cause long-term side effects down the road?

Don't know.

But after a thoughtful few moments, he recommended I stop. My original cancer had been middle of the road; stage IIB, only moderately aggressive. And I'd developed osteopenia (bone loss) as a result of taking Arimidex. Stopping the Arimidex would (hopefully) stop the bone loss.

He made a judgment call based on his years of experience as an oncologist, and the limited information about Arimidex we had. And I agreed; I happily threw away the Arimidex bottle with its remaining three pills, and have been recurrence-free thus far.

I hope, down the road, that clinical trials will show the optimal duration for taking Arimidex (or any AI) is 5 years. But if it turns out 10 years is better, well, what the heck. My oncologist and I did the best we could with the information we had.

If you're facing the decision of whether to continue your AI, there's no data to help you. You'll have to rely on how you're handling any side effects now; how aggressive your original cancer was; and how comfortable you are bucking current ASCO (American Society of Clinical Oncology) guidelines, which call for 5 years of AI treatment.

Do the best you can, with the information you have. That's all any of us can do.


Burstein, H. G. et. al. (2010, July 12). American Society of Clinical Oncology clinical practice guideline: update on adjuvant endocrine therapy for women with hormone receptor-positive breast cancer. Retrieved from

Aromatase inhibitors as adjuvant therapy for breast cancer. (2011, February 4). Retrieved from