Overdiagnosis in Thyroid Cancer: The Case for Observation

An Interview With R. Michael Tuttle, MD

R. Michael Tuttle, MD; Gabriel Miller


August 25, 2015

Editor's Note:
Thyroid cancer represents a recent and controversial example of potential overdiagnosis in cancer. One of the most frequent arguments made in this debate is that thyroid cancer incidence has skyrocketed in recent years, but mortality has remained largely unchanged.

For example, the incidence of papillary thyroid cancer has doubled over the past decade; however, before 2002, it took 30 years for its incidence to increase a similar level.[1] Thyroid cancer now ranks as the ninth most common cancer in the United States and the fifth most common cancer among women, but thyroid cancer has the lowest mortality rate among the top 10 cancers.[1]

Treatment for the disease is not without the potential for complications and, regardless, comes with lifelong effects. Only recently has the field sought to actively counter the risks of overdiagnosis and overtreatment with a more conservative approach.

Medscape spoke with R. Michael Tuttle, MD, an endocrinologist at Memorial Sloan Kettering Cancer Center who specializes in advanced thyroid cancer, about overdiagnosis of the disease, how he advises patients who present with smaller thyroid nodules, and the questions the field needs to answer to move forward. Dr Tuttle served previously as chair of the National Comprehensive Cancer Network thyroid cancer guidelines committee and currently is a member of the American Thyroid Association (ATA) guidelines committee.

Medscape: Are we truly seeing overdiagnosis of thyroid cancer at the moment?

Dr Tuttle: We are clearly overdiagnosing thyroid cancer. If you look at the rates of thyroid cancer in the United States, particularly of the subcentimeter papillary microcarcinomas, it is going right through the roof. And it has become apparent that much of what we are diagnosing is not clinically relevant at all—that the vast majority of these thyroid nodules would have stayed small for 100 years without any intervention. We really have to question whether we need to diagnosis every 2-mm thyroid cancer.

Medscape: One of the arguments used to support the case of overdiagnosis is that there has been quite an increase in diagnosis but no change in survival. That is, if you look at the history of the incidence of the disease, this is a recent increase that has happened over the past 10 or 15 years. What have been the primary causes of this increase?

Dr Tuttle: I always hear overdiagnosis advocates say that there is no change in mortality of thyroid cancer. The reality is that the disease-specific mortality of small papillary thyroid cancer is so low anyway that it would be hard to show a difference.

That being said, there is no question that there are a number of good studies that show a relationship among the number of ultrasounds in the community, the number of endocrinologists in the community, and the likelihood of being diagnosed with thyroid cancer. There is no question that overuse of ultrasound and fine-needle aspiration biopsy is leading to this overdiagnosis of early, small-nodule disease.

Ultrasounds are often being used for entirely inappropriate reasons. I saw somebody today whose hair was getting thin, and her doctor decided to do an ultrasound to check her thyroid. Of course, they found a 5-mm papillary thyroid cancer that we are now going to follow with observation. In other cases, patients get carotid ultrasound, or chest or neck CT, and we find little changes in the thyroid, which usually leads to an ultrasound.

Because at least 50% of the adult population has thyroid nodules and as many as 10%-15% have small thyroid cancers, it is not surprising that ultrasonographic thyroid exams in asymptomatic people will identify lots of thyroid nodules and small thyroid cancers.

Furthermore, we can now biopsy thyroid nodules as small as 2 and 3 mm. When I was a fellow, we would have never biopsied something smaller than 10 or 15 mm.

So we have excessive use of thyroid ultrasound that is being ordered for inappropriate reasons, combined with the ability to biopsy, very safely and very well, these really teeny tiny things. That sets the stage for a huge increase in small-volume thyroid cancer.

Medscape: These are primarily incidental findings as a result of ultrasound for other causes?

Dr Tuttle: Most the time, they are. Thyroid cancer can sometimes be found by physical examination either as a thyroid nodule or as metastatic lymph nodes in the neck. Clearly, this is clinically significant thyroid cancer that needs to be diagnosed and treated.

However, even incidental findings can be important. I have seen patients with thyroid nodules as large as 2-3 cm that cannot be palpated on examination. So not every incidental finding should be ignored.

By definition, these very small thyroid cancers are either being diagnosed because the ultrasound is being done for the wrong reasons, or to follow up a CT, MRI, or something else that was done for the right reason but accidently found a small thyroid nodule.

Medscape: What role do radiologists and oncologists play in overdiagnosis and presumably overtreatment, respectively, of thyroid cancer?

Dr Tuttle: We are all to blame. Clearly, radiologists do not have much of a choice. If I request an ultrasound, then they are going to have to describe a small nodule—they can't ignore what they see.

Endocrinologists are the same way. If they see a nodule, then they usually feel that they should know what it is; they feel they should biopsy it. Patients are the same way. Common sense says that the earlier that you find a thyroid cancer, the better, whether you are a radiologist, endocrinologist, oncologist, or a patient. Common sense says that if it is thyroid cancer, you should find that out.

However, the new guidelines from the ATA, when they come out this fall, are going to specifically recommend against biopsying subcentimeter nodules even if they look suspicious. It is the first time a group has made that strong a recommendation saying that even if a small thyroid nodule looks very suspicious for thyroid cancer, if it looks like it is confined to the thyroid, then the right thing to do is to repeat the ultrasound in 6-12 months, not to stick a needle into it. Once you biopsy it, it is really hard for patients not to do anything about it.

The real issue is first, who is ordering that ultrasound to begin with, and second, who is insisting on having these small things biopsied. I think this is being driven by primary healthcare providers and endocrinologists and less so by radiologists and oncologists.

Medscape: The guidelines are indeed becoming clearer about the subcentimeter nodules, with societies and organizations beginning to recognize nodules that should be left alone. Is there an issue with getting this guidance to primary care doctors and endocrinologists?

Dr Tuttle: Absolutely. We are fighting common sense. Nobody ever taught clinicians they were supposed to biopsy 5-mm thyroid nodules. You can't find that in any of the guidelines, yet you just kind of know that is the right thing to do. So we are fighting against what is a kind of common sense, and the recognition that we don't need to biopsy every subcentimeter nodule has really only gathered hold in the literature in the past several years.

The recommendations from the ATA that specifically say not to biopsy nodules smaller than 1 cm are going to be pretty controversial when they come out, and there are going to be a lot of people who disagree with them. We are only at the edge of the very senior thyroid specialists in the United States recognizing that immediate diagnosis and therapy are not required for every small papillary thyroid cancer. The challenge will be translating this information to endocrinologists, primary care doctors, family practice physicians, and patients and their families.

Medscape: You used the phrase "common sense" to describe diagnosis, meaning that common sense says to look at a thyroid nodule early and find out what it is. However, there has also been this rise in ultrasound that has contributed to overdiagnosis. How does the US healthcare system, the reimbursement system, incentivize diagnosis and treatment of thyroid cancer? Does that contribute in any way?

Dr Tuttle: No, honestly, I don't think so. You don't make a lot of money doing ultrasounds and biopsies. But there is no question that you get paid if you do that, and surgeons are going to be paid to do surgery.

It just doesn't feel to me like monetary incentives are what is driving it. What's driving it is the desire to make a diagnosis of a cancer and to treat cancer. When I talk to doctors who are doing early diagnosis and treatment, they truly believe in their heart of hearts that they did a very good thing by finding a 5-mm papillary thyroid cancer.

I find that when I give lectures and talk about these issues, the people who give me grief are not worried about incentive or cost or money. They think that by not doing something, we are preventing them from curing some small cancer.

You can't totally take the incentive piece out of it, but that truly doesn't feel like what is driving it to me.

Medscape: How should clinicians balance the risk/benefit ratio of biopsy and treatment when they find a nodule? The examples we have been discussing are very small—5 mm. However what if they are a little bit larger and more on the edge of clinical relevance, if you will? What guidance do you give to clinicians in that situation?

Dr Tuttle: First of all, you have to admit that observation might be a reasonable option because up until now, that has never been recognized as a reasonable option. Up until this point, if you had thyroid cancer, you automatically got treated for it because there was no other alternative. If you have cancer, you just have to accept whatever the complications are of treating that cancer.

So the first step is for clinicians to take a step back and say, "Okay, is this one of those very low-risk thyroid cancers that could be followed with observation?" That is the first step.

Then if you intellectually say, "Well, this is a small one that could be followed with observation," you look at the risks and benefits of observation and the risks and benefits of surgery.

With observation, the benefits are that you avoid a surgery and you may not need to be on thyroid hormone pills. The risks are that the thyroid cancer could grow or spread.

With surgery, the benefits are that you've taken the thyroid out and the cancer is out of your body in terms of where we saw it, but the risks are that you could damage a nerve, the vocal cord, or the parathyroid.

What I find is that some patients perceive these risks differently, or more, than others. I treat professional voice users, and if we damage their voice at all, then they can't sing on Broadway, lecture, teach, or be a lawyer. So they're much more willing to accept the risks of observation. Others can't stand the thought of having a small thyroid cancer in their body, so they won't accept the risks of observation.

It really has to be a conscious decision, where you first recognize that there are in fact two options for some of these patients and then carefully go through the risks and benefits of both options with each patient.

Medscape: You are involved in a National Cancer Institute-funded study looking at how to separate the different subtypes of low-risk thyroid cancers. What do we need to learn in order to separate indolent from potentially aggressive small thyroid cancers? Similarly, do valid strategies to do this exist that we are already aware of, but simply haven't been disseminated into clinical practice yet?

Dr Tuttle: There are two levels of answers. One is from the studies by Ito and colleagues[2] and Sugitani and colleagues,[3] which are the Japanese studies of observation. They have showed some clinical factors that help predict which of the thyroid cancers are going to be indolent and which are going to be more aggressive.

The indolent tumors tended to be small tumors that developed in older patients. Younger patients are slightly more likely to see disease progression during observation, as are patients who get pregnant during observation. Patients who demonstrate ultrasonographic evidence of growth of the thyroid cancer nodule, extrathyroidal extension, or cervical lymph node metastases usually continue to progress and are not good candidates for observation.

These are some clinical features that help you know whether this would be a thyroid nodule you could watch or not. Obviously, they are not good enough, and we've been interested in trying to find a molecular signature. When we have one of these small thyroid cancers, what we would love is to be able to check 50, 100 or 200 genes that would then tell us which ones are going to grow and which ones are not. The simple gene panels we have right now don't do it. Everybody knows about BRAF, but it turns out that about one half of these papillary microcarcinomas have BRAF and yet 90% of them don't grow. It has to be more complex than that.

We are doing a study now where we have about 250 low-risk papillary thyroid cancers that we're following with observation, where we haven't done surgery. In the few that grow, the patients have agreed to give us a piece of their tumor when they go to surgery, and we are going to compare the genes in the people whose tumors grew with the genes in the people whose tumors didn't grow. Yuri Nikiforov, MD, PhD, in Pittsburgh is doing a similar thing, but he is trying to use molecular analysis of fine-needle aspiration biopsy samples in combination with clinical and histologic features to predict clinical outcomes in small thyroid cancers.

We hope in a few years to have molecular signatures that help predict whether a thyroid nodule is going to grow, spread, or be stable. Certainly, that is ready for implementation into clinical practice right now.

Medscape: What is your main message about papillary microcarcinomas?

Dr Tuttle: My main message is that for properly selected patients, an observational approach to initial management of papillary thyroid microcarcinoma is a viable option to immediate surgery. What is critical is that it is okay for me to be wrong for 6 months. If we recommend not to operate on someone, we don't say go away and come back in 25 years. We say, "We don't think you need the surgery now, and we'll do an ultrasound every 6 months for about 2 years. This is papillary cancer. This is not melanoma or pancreatic cancer. In the few that grow, they grow over a year or two, and even if they grow, our surgical option is going to be just as curative as if we do it now."

When I lecture, I tell people I don't mind observing patients even though I know that the tumor will grow in about 10% of them and 2%-3% of them are going to get lymph node metastasis, because our treatments are going to be just as effective then as if I used them right now. What I try to convince them of is that we are not choosing between "yes" to surgery and "no" to surgery. We are choosing between surgery now vs watch for a while and see whether you really need surgery. It's hard to argue that I hurt anybody by watching them for 6 months in low-risk thyroid cancer.


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.