High Thyroid Cancer Mortality Rate in Japan: A Result of Nonaggressive Treatment Strategy,or Just an Aging Population?

Dear Editor:

The Surveillance, Epidemiology, and End Results (SEER) database for the United States (1) and cancer statistics for Japan (2) show that the mortality rate from thyroid cancer in the United States recently was 0.5 per 100,000 men and women per year, whereas this rate in Japan was higher at 1.4. But there are caveats related to age.

Although this difference may seem small, it suggests that each year there are about the same number of thyroid cancer–specific deaths (about 2000) in Japan as in the United States, even though the national population of Japan is much smaller. What may explain this quite large difference in mortality rates?

The Japanese have more iodine in their diet, and the national population is more aged than in the United States, but there is no evidence that the prevalence rates of the various main cancer types differ between the two countries, or that differentiated thyroid cancers have different mutations, or somehow are different diseases in the two regions. Of course, for decades the Japanese have managed differentiated thyroid cancer tumors in the 1 cm to 3 cm diameter range less aggressively than in the United States, but could this really be responsible for the death rate in Japan being more than double the rate in the United States? As compared with the United States, the three main clinical management differences in Japan during the past 30 to 40 years have been (a) reluctance to use radioiodine for remnant ablation after thyroidectomy, (b) a preference for lobectomy rather than thyroidectomy, and (c) a preference not to prescribe chronic thyroxine therapy for lobectomy patients, but instead to allow the patients to live with serum thyrotropin (TSH) levels that fall at the high end of the normal range. I had been particularly surprised a few years ago to learn about this preference not to prescribe thyroxine, but it has to do with prior legal limits to refills on prescriptions, the desire to avoid an “unnatural” treatment, and the presumed inconvenience and expense of taking a daily medication for decades.

Let me quickly add that my observations are not intended to represent any criticism of the clinical care being given by my highly skilled Japanese colleagues. I was prompted to write this letter by the recent appearance in Thyroid of the manuscript by Hall et al. (3). These authors found that regions in Canada with more aggressive or less aggressive thyroid cancer treatment regimens appeared to have similar clinical outcomes after follow-up of about 15 years. Of note, the clinical outcomes of patients in Japan do not differ from patients in the United States for the first 10 or 15 years as well (2), but the older Japanese citizens seem to have increasing problems with serious thyroid cancers. We are all aware of the trend in United States to recommend less aggressive treatment for small differentiated thyroid cancers; therefore, any signal that this treatment strategy may be associated with higher disease-specific mortality rates should cause some reflection. But an association does not prove causality. In fact, Japan has another important factor potentially at work—a more aged population than is present in the United States.

It is striking that in Japan during 2015, 68.6% of the 1729 deaths due to thyroid cancer occurred in persons aged 75 years and above. Specifically, of the 1729 deaths, the numbers in age ranges 75–79 years, 80–84 years, and 84+ years were 259, 342, and 584 deaths, respectively (2). The Japanese mortality rates for these three age ranges were, respectively, 4.1, 6.8, and 11.6 per 100,000 men and women within each age range (2). SEER has a summary graphic that presents the U.S. data in a somewhat different format covering age distribution of thyroid cancer deaths, both sexes and all races, for combined years 2010 through 2014 (1). In the United States, the age range 75–84 years contained 27.2% of thyroid cancer deaths, and there was a decline in the 84+ group to 19.0% of total thyroid cancer deaths. One expert has told me that the “age-adjusted” disease-specific mortality rate in Japan is not higher than in the United States, but another expert I spoke with disagrees. I am not an epidemiologist, but I write this letter to see if experts might have an interest in further analyses.

It was my privilege to visit Japan several times during the past decade to work with thyroid cancer experts there who were completing a small pivotal clinical study (n = 10) intended to secure registration of recombinant human TSH (rhTSH) for use in patients in Japan. Nearly all of the 10 patients had been diagnosed and initially treated nonaggressively for differentiated thyroid cancer, on average 15.1 years prior to study enrollment. Treatments had been viewed initially as successful, but the patients then developed distant metastases of thyroid cancer years later. These patients underwent completion thyroidectomy and then were entered into the study. The goal was to assess the ability of rhTSH and whole body scanning to detect distant metastases. This “diagnostic” study was successful and was published in the Japanese literature by Konishi et al. (4). I presented the study data orally to the Japanese regulators along with other supporting information about rhTSH, and our team successfully achieved registration of the drug. I did learn from my visits that most patients newly diagnosed with differentiated thyroid cancer in Japan did well for many decades after lobectomy and no radioiodine, but that “a few unlucky patients,” as they were described to me, presented with distant metastases about 15 to 20 years later. At that point, completion thyroidectomy and high activity radioiodine were recommended, although their prognosis was grim. I want to add that the drug Thyrogen™ does not have formal regulatory approval for use in combination with radioiodine for the treatment of distant metastases, and a number of serious complications related to tumor swelling could occur if the drug were used in this clinical setting.

It appears that the healthy lifestyle of Japanese citizens allows many to reach age 80 and beyond, setting them up to present for the first time with thyroid cancer at an advanced age. But one also wonders whether less aggressive treatments of middle-aged Japanese patients may be putting the older cancer survivors at greater risk of thyroid cancer death, perhaps as additional mutations in tumor cells occur, or as the patients' immune systems weaken with age. It does seem to me that if specialists in the United States adopt less aggressive treatment paradigms more widely, it may mean that thyroid cancer survivors will need many years of neck ultrasounds, serum thyroglobulin evaluations, and perhaps other testing to watch for tumor growth. Chronic prescription of thyroxine to lobectomy patients and monitoring of their TSH levels also might be indicated, although recent patient series of thyroxine replacement after lobectomy have not demonstrated clear benefit. Yet in patients with advanced differentiated thyroid cancer, keeping serum TSH in the low normal range did prove helpful.

In sum, in this brief letter I want to share my personal concern and confusion about this area of thyroid cancer management and outcomes, and I ask whether further discussion would be useful, as there are decades of high-quality Japanese data available. My advice is to focus on deaths. Experts may be talking past one another, however, since what may matter is the “number needed to treat to achieve significant benefit.” Consider a thousand 45-year-old patients who each have a 2.2 cm papillary thyroid cancer with two 0.9 cm tumor-involved central lymph nodes (without capsular invasion of the tumor or extranodal extension). How many of the 1000 patients would be significantly helped (or harmed) by near-total thyroidectomy and 30 mCi radioiodine remnant ablation? If only three of the 1000 avoid later onset of distant metastases (need to treat 333 to help one), is that worth it? What if 30 patients are helped? And the answer to the question about risk/benefit likely must factor in local variables, such as available surgical expertise. But let me raise a specific number: what if one patient with “low-risk” differentiated thyroid cancer received significant benefit for every 153 such patients who received near-total thyroidectomy plus 30 mCi radioiodine remnant ablation? We can make an analogy about this quantity.

I recently turned 66 and went through the inconvenience and expense of my scheduled screening colonoscopy. Every year several million Americans have such screening colonoscopies, but studies show a favorable risk:benefit ratio and cost effectiveness. Of course, there is not an exact analogy between removing small cancerous colon polyps and small cancerous thyroid tumors. Imperiale et al. found that of 9989 evaluable persons at average risk for colon cancer who underwent screening colonoscopy, cancers were found in 65 (0.7%), although possibly precancerous lesions were found in some others (5). With a focus on actual cancers, one person for every 153 persons screened significantly benefited using colonoscopy, and this rate was deemed a success.

After hearing many quality presentations about the merits of less aggressive treatment of low-risk differentiated thyroid cancer, I still have an uneasy feeling about this trend. But I plan to continue to be educated by the experience of my skilled colleagues.