Thyroid Cancer in Fukushima Children: When the Language and Information Gaps Mislead

With the year 2016 marking the passage of five years since the Fukushima nuclear accident, many writings—articles, editorials, academic papers—have been released reflecting on the first five years after the accident. Some of the writings address a psychosocial aspect of the accident such as “problems” caused by the stress of evacuation and the “unwarranted” fear of radiation, dismissing the potential health effects of radiation exposure, even ignoring the science. Others focus on the alleged withholding of medical data by authorities, speculating on the health effects of the Fukushima accident reaching even the United States. 

Official data and information available in English are often limited and biased. Transparency and impartiality of such information, released by the government and international agencies, can be influenced by ulterior motives other than public health protection. However, without a fluency in Japanese and an ability to navigate through and comprehend the mass of official and unofficial information only available in Japanese, it may not become obvious that the transmission of accurate information is indeed hindered by the language barrier.

Furthermore, followers of numerous government committee meetings regarding the health effects of the Fukushima nuclear accidents—most of them live streamed on the Internet—have witnessed a systematic underestimation of health effects due to low-dose radiation exposure, with the claim of the outdated and unscientific 100 mSv threshold discourse. Despite concerns from local medical associations, potential health effects in prefectures adjacent to Fukushima Prefecture were dismissed, as if the radioactive plume was blocked by an invisible wall at prefectural borders. This is a far cry from the precautionary principle that should be in place for the protection of public health. 

Consider the Japanese government’s haste to return evacuees to their still contaminated hometowns. This must be done so things appear “back to normal” for the purpose of recovery (mostly economic), even though it is clearly impossible to decontaminate a whole community in a natural setting of mountains and forests. Radiation doses of returned residents are to be monitored to keep an additional exposure dose below the regulatory limit. (But how good is it to know what your exposure dose is after the fact?) 

In essence, the health effects by the Fukushima nuclear accident are being maximally minimized. 

One of the most controversial topics about the health effects of the Fukushima nuclear accident is the thyroid cancer cases detected in Fukushima children as a result of the thyroid ultrasound screening. Most of the English writings on this topic accept, at face value, certain claims made by Fukushima Medical University as well as Japanese government officials in order to dismiss any connection between the Fukushima thyroid cancer cases and radiation. 

Below, some items in the March 25, 2016 editorial in Chicago Tribune, “The children of Fukushima: When medical tests mislead,” are addressed to point out the misleading information that is widespread even amongst the academic circle. 

1. There is no regional difference of thyroid cancer occurrence.

The March 25, 2016 Chicago Tribune editorial states:

“Children living closer to the accident in areas of greatest contamination had no greater rate of early cancer than those living farther away.”

This essentially refers to the lack of dose response, but it might depend on how the prefecture is divided into regions. 

According to the official data by Fukushima Medical University (FMU) and Fukushima Prefecture in the final report of the first round screening [1], no regional difference was reported based on the comparison amongst 4 regions—one region including 13 municipalities with the highest dose and the evacuation zone, and three other geographically-divided regions. However, topography can vary even within the same geographical region, potentially affecting the flow of the radioactive plume. In other words, regional divisions like this might mask critical differences.

On the other hand, the biggest surprise in the official comparison was the Aizu region in western part of Fukushima Prefecture where the prevalence rate of 32.6 per 100,000 was very close to the prevalence rate in the highest dose area, 33.5 per 100,000. (See the second row from the bottom in Table 9). 

The Aizu region supposedly received very little contamination owing to the mountain range to its east, and the officials seem to attribute the higher-than-expected prevalence to the screening activity. According to this logic, other uncontaminated areas of Japan might have similar prevalence. In fact, thyroid ultrasound screening of 4,365 subjects in three other prefectures—Yamanashi, Aomori and Nagasaki—called the three-prefecture study [2] in short, revealed about the same frequency of ultrasound findings such as cysts and nodules as Fukushima, and one case of papillary thyroid cancer was diagnosed in the follow-up study [3].  Very little discussed are the limitations of this study, such as a small sample size leading to a bigger uncertainty, and different age and gender distributions (no age 0-2 and more female adolescents) than Fukushima.

If the three-prefecture study is indeed a proof of screening effects in Fukushima, then it can be deduced that similar prevalence might be found in the rest of Japan. Judging from the clinical information provided by Shinichi Suzuki, a thyroid surgeon at FMU, the vast majority of the operated cases in Fukushima had surgical indications, i.e. required surgery, such as the metastasis and proximity to vital structures, even in the absence of subjective symptoms at the time of diagnosis. Then, it logically follows that there may be other such cases outside Fukushima. However, no attempt is made to conduct a nationwide survey for fear of causing detriments to asymptomatic individuals through unnecessary medical tests and treatments.

In October 2015, the Okayama University team headed by an environmental epidemiologist and public health specialist, Toshihide Tsuda, published in Epidemiology the first epidemiological analysis of the publicly available Fukushima thyroid cancer data [4]. Tsuda et al. first divided the entire Fukushima Prefecture into 3 areas by the year of screening and subdivided them into 9 districts based on the population size, enabling a statistical analysis. 

For external comparison, they calculated for each area/district the incidence rate ratio from the prevalence rate assuming a latency period of 4 years. The calculated incident rate ratios showed thyroid cancer occurrence in some regions were 30 to 50 times the expected rate. Tsuda et al. concluded the excess occurrence was beyond what could be explained by screening effect. Internal comparison did not appear to show dose response amongst the 3 regions by the screening year, which supposedly represented the degree of contamination. 

However, when the data (updated to June 30, 2015) was adjusted for “latency” or the length of time elapsed between the accident and time of screening in their response to letters to the editor [5], a dose response tendency showed up with higher prevalent odds ratios in districts closer to the Fukushima Daiichi nuclear power plant.

Other points to keep in mind include the fact the dose estimation is far from perfect with insufficient direct measurements of the thyroid dose and that the soil contamination does not necessarily reflect the amount of radioactive plume that was in the air at the time of exposure, and that the cesium soil deposition does not necessarily correlate with the amount of radioactive iodine that might have been inhaled or ingested.

A small number of measurements conducted in some 1080+ children aged 0-15 from three municipalities—small considering there were over 360,000 children aged 18 or younger in Fukushima Prefecture at the time of the accident—were ladened with technical issues such as high background radiation levels. The March 30, 2011 internal document, a correspondence from the Emergency Response Center to the Nuclear Safety Commission [6], states how 11 highest readings, taken in Iwaki City, were adjusted to yield lower thyroid equivalent doses as shown in the table below. 

From left of the table, column 1 shows measurement dates of March 26 or March 27, 2011. Columns 2 and 3, blacked out for privacy protection on the original document, correspond to the identification number and age of the subject, respectively. Column 4 shows the actual reading ranging from 0.22 μSv/h to 0.27 μSv/h. Columns 5 and 6 correspond to the net measurement after subtraction of the average air dose in μSv/h as the background level in and the resultant thyroid equivalent dose in mSv, respectively. Columns 7 and 8 correspond to the net measurement in μSv/h after subtraction of the measurement at the clothed shoulder as the individual background level and the resultant  thyroid equivalent doses in mSv, respectively. 

For example, in the first subject with the actual measurement of 0.22 μSv/h, the net measurement went down from 0.05 μSv/h to 0.04 μSv/h and the thyroid equivalent dose went down from 19 mSv to 7.1 mSv. For subject 8 with the highest reading of 0.24 μSv/h who had the highest thyroid equivalent dose of 43 mSv before the adjustment, using the individual background level taken at the clothed shoulder reduced the thyroid dose to 25 mSv. 

The problem with using the measurement taken at the clothed shoulder is the potential contamination on clothing might lead to a smaller net measurement. For instance, if the skin and the clothing have the same amount of radioactive contamination, subtracting the measurement of clothing from the measurement of the thyroid will yield the net measurement of zero, but this may not reflect the actual thyroid measurement. 

Despite its limitations and the potential underestimation, the so-called 1080 survey data is considered to represent the exposure dose of Fukushima children, and it seems to be widely accepted by overseas researchers who rely on information published in peer-reviewed English language journals that the maximum thyroid dose in Fukushima children was 35 mSv. Yet there are many other children who might have been exposed to higher doses but never had thyroid doses measured. For instance, children from non-evacuation zone led regular daily activities without any protective measures in the absence of any warnings about the approaching radioactive plume. Many were outside unprotected, and there is no way to accurately estimate the exposure doses for them.

In addition, the Ministry of the Environment expert committee chaired by Shigenobu Nagataki heard presentations by experts who actually conducted the 1080 survey who acknowledged that they thought the survey was only preliminary and more detailed measurements would be conducted if necessary. In the end, officials never conducted a more detailed thyroid survey measurement in a boy from Iwaki City who registered the highest thyroid measurement, 35 mSv, for fear of “scaring the family and the community” upon a recommendation [7] of Yoshiharu Yonekura, the chair of the National Institute of Radiological Sciences as well as the current chair of UNSCEAR. (Incidentally, Iwaki City, not one of the evacuated municipalities, was hit by the radioactive plume when the wind turned south, but the lack of precipitation afterwards prevented a heavy ground deposition of radioactive substances as in Iitate Village).  

Maintaining a sense of security and preempting panic and mass chaos took precedence over accurate recording/reporting of exposure doses. 

2. Age distribution

The March 25, 2016 Chicago Tribune editorial states:

“Younger children and infants, whose thyroids are more likely to be affected than those of older children, did not show an expected higher rate of abnormal findings.”

It is well known that beginning about 4 years after the Chernobyl accident, an increased number of thyroid cancer cases were detected in children in Russia, Belarus and Ukraine who were infants at the time of the accident. In the first 3 years after the Fukushima accident, the youngest cancer case was 6 years old at the time of the accident according to the data published so far. This is depicted in the age distribution graph in a letter to the editor of Thyroid [8].

The top panel shows the number of thyroid cancer cases by age in Ukraine in two different time periods: 1986-1989 (black) and 1990-1993 (dark gray). The 1986-1989 age distribution resembles the graph in the bottom panel showing the first three years in Fukushima (light gray). 

Japanese authorities, notably Shunichi Yamashita, have referred to this resemblance to make certain points: 1) Fukushima’s age distribution in the first three years resembles the first three years in Chernobyl during the latency period when radiation-induced thyroid cancer was still growing, and thus the Fukushima cases are not related to radiation; 2) Fukushima’s thyroid cancer cases are not related to radiation exposure since they do not include younger children under age 4 or 5. 

It might be this second point that is addressed by author(s) of the Chicago Tribune editorial. However, this cannot be a valid claim since a large number of cancer cases in younger children under 4 or 5 were not seen until 4 to 5 years after the Chernobyl accident, and it should not be compared to the first 3 years in Fukushima. Yet this misguided direct comparison is done by none other than the authoritative researchers in radiation-induced thyroid cancer such as Williams [9] and Wakeford [10].

(From: Williams D, Thyroid Growth and Cancer. Eur Thyroid J 2015;4:164-173)

(Wakeford J. Radiol. Prot.  36  (2016) E1)

It remains to be seen whether a similar increase in thyroid cancer cases in younger children will be seen in Fukushima from the fourth year onward, when the complete results of the ongoing second round screening covering the fourth and fifth post-accident years become available, probably later in 2016.

3. Latency issues

The Chicago Tribune editorial states:

“Some cancers were observed less than a year after the meltdown. It generally takes years for thyroid cancer to develop after radiation exposure.”

For clarification, the first case suspicious for thyroid cancer in Fukushima children was reported in September 2012, or 18 months after the accident, not less than a year. 

A traditional view of carcinogenesis is a multi-stage process of mutations transforming normal cells into cancer cells through initiation, promotion, and malignant transformation , and it is considered to take a long time. Radiation-induced thyroid cancer in children is thought to take several years to develop after DNA damage is “initiated” by radiation,  which might explain the increase in thyroid cancer cases 4-5 years after the Chernobyl accident in children who were younger than age 5 at the time of the accident.

On the other hand, there were higher-than-expected number of thyroid cancer cases diagnosed in Belarus the first year after the 1986 Chernobyl accident [12,13].

As a complete carcinogen, radiation can affect all stages of carcinogenesis. Radiation can also affect the so-called “cancer niche,” or the tumor microenvironment necessary for clinical development of cancer [14]. In other words, radiation might facilitate the growth of pre-existing precancerous cells or latent cancers through promotion as well as creation and/or expansion of the cancer niche.

It might then be feasible to consider that some of these early cases manifested because exposure to radiation affected the growth of pre-malignant cells or pre-existing cancer. This might not be considered “radiation-induced” in the field of radiobiology, but it can be considered as the effect of radiation nonetheless.

4. Overdiagnosis?

Yet, the reason for detection of so many thyroid cancer cases in the first 3-4 years after the Fukushima accident is not completely clear. Is it because of the screening activity alone as claimed by the officials? Or is it due to radiation exposure? Or is it for some other causes?  

The Chicago Tribune editorial concludes that thyroid ultrasound screening in Fukushima is causing overdiagnosis and suggests overtreatment might be damaging the Fukushima children, but clearly author(s) did not review the clinical details of the operated cases. 

This is not surprising because clinical and pathological information of the operated cases has not been officially released in English and thus not readily available. Pieces of information have trickled out, mostly in Japanses, as abstracts for presentations at Japanese academic conferences and two reports submitted to the Thyroid Examination Evaluation Subcommittee of the Fukushima Health Management Survey Oversight Committee (see the November 29, 2014 post, the June 23, 2015 post, and the September 10, 2015 post for details).

It is important to understand that clinical and pathological information released by FMU is not comprehensive enough to allow for any meaningful analysis of the surgical cases. However, even with such limited information, it is clear that  surgeries were medically indicated for the vast majority of the cases for reasons such as lymph node and/or distant metastases and close proximity of the tumor to vital structures such as the vocal cord, trachea, or recurrent laryngeal nerve. Pathological analysis of the surgical specimen revealed 39% had mild extrathyroidal extension (cancer cells spreading beyond the covering of the thyroid gland), and 74% had lymph node metastasis. In order to preserve thyroid hormone secretion, only half of the thyroid gland was removed in 94% of the operated cases. (At the press conference of a newly launched “311 Thyroid Cancer Family Group [15],” a physician advisor to the group revealed there were a few cases of recurrence, but this information has been neither released nor confirmed by FMU). 

Furthermore, the interim result of the second round screening shows 40 (80%) of 51 cases suspected or confirmed of thyroid cancer had no ultrasound findings of precancerous tumor in the first round screening—in fact, 25 cases (50%) had no ultrasound findings at all. In other words, for the majority of the cancer cases diagnosed in the second round screening, cancer grew in two to two and a half years since the first round screening.  

Thus the clinical information suggests aggressiveness and fast progression of the thyroid cancer being diagnosed in Fukushima children. (Incidentally, the April 14, 2016 New York Times article [16] reports a type of thyroid cancer, encapsulated follicular variant of papillary thyroid cancer, is reclassified as a benign thyroid tumor [17]. It has been reported that 3 of 87 thyroid cancer cases operated by the end of 2014 were classified as the follicular variant of papillary thyroid cancer [18]. However, there is no sufficient information regarding whether these 3 cases were encapsulated or not).

Two groups of Japanese researchers did calculations to conclude the thyroid cancer occurrence is about 20-60 times the expected rate. Both groups seem to agree this is more than expected from the screening effect—detection of “quiet” and asymptomatic thyroid cancer due to the screening activity—alone.

Tsuda et al., aforementioned in section 1, claim that the excess cases might reflect some screening effect but the radiation effect should also be considered, pointing out the potentially higher exposure doses and the clinical features of the cancer cases.

The study by Tsuda et al. drew a number of criticisms domestically and internationally, mainly for comparing prevalence obtained from screening with incidence of clinically diagnosed cases. Local newspapers in Fukushima Prefecture never covered the release of the Tsuda et al. study. However, a second group of researchers conducted a similar comparison using different methods without such overt criticisms. 

In November 2014, Shoichiro Tsugane, a National Cancer Center epidemiologist and a member of the Fukushima Health Management Survey as well as the Thyroid Examination Evaluation Subcommittee, presented to the Subcommittee a document showing how many cases of thyroid cancer would have been diagnosed in Fukushima Prefecture in 2010, the year before the accident, calculated from the national annual incidence rate [19, 20]. He concluded the observed cancer cases were 61 times the expected and attributed the excess to overdiagnosis. The same estimate was also published in English [21]. In 2015, a related paper was published by Tsugane’s team with updated data, showing the observed rate was about 30 times the expected rate and attributing this increase to overdiagnosis due to screening [22]. Their estimate is included in the Interim Summary by the Fukushima Health Management Survey Oversight Committee [23] as the expression, “several tens of time larger than expected.” 

From published interviews of Tsugane, it is clear he is not considering clinical details of the surgical cases [24] in claiming overdiagnosis which means diagnosis of cancer that does not necessarily need medical intervention. 


When writings such as the March 25, 2016 Chicago Tribune editorial mislead readers on facts of the issue, it is damaging to Fukushima people whose plight is lessened and disregarded. But the core issue is the release of biased information by the authorities, especially in English. This gap in information only adds to the pre-existing language barrier, distorting information transmission.

It is essential for every one of us to be well informed and develop critical thinking in order to counter the one-sided flow of distorted information. The recently established family support group for thyroid cancer patients, “311 Thyroid Cancer Family Group,” will provide the platform for the patients and families to voice their concerns. What is urgently called for is a comprehensive evaluation of data in a transparent and unbiased manner by third-party expert groups, such as the International Society for Environmental Epidemiologists [25].


Fukushima Thyroid Examination February 2020: 186 Surgically Confirmed as Thyroid Cancer Among 237 Cytology Suspected Cases

Highlights   The third round: 1 case newly diagnosed as suspicious or malignant, and 5 new cases surgically confirmed.  The fourth ...