Fukushima Thyroid Examination Fact Sheet: March 2016


Fukushima Thyroid Examination Fact Sheet:  March 2016

This fact sheet was compiled mostly from the official documents and data, some available only in Japanese, for the purpose of reviewing the first four and a half years of the post-accident thyroid ultrasound screening in Fukushima Prefecture. It is intended as a common knowledge base of the history, basic principle and current status of the largest thyroid ultrasound screening undertaken with controversial results. (For the shorter version of the fact sheet, please refer to this).



Introduction

On October 9, 2011, Fukushima Prefecture began the Thyroid Ultrasound Examination (TUE) on about 360,000 residents who were age 18 or younger at the time of the triple disaster of the earthquake, tsunami, and nuclear accident on March 11, 2011. As the exposure to radioactive iodine dramatically increased the incidence of pediatric thyroid cancer cases after the 1986 Chernobyl nuclear accident, TUE was implemented to monitor the exposed children in Fukushima Prefecture. The majority of Fukushima residents did not receive stable iodine for protection of their thyroid glands.

TUE is part of the Fukushima Health Management Survey (FHMS) [1],  consisting of Basic Survey for external radiation exposure dose for the first four post-accident months estimated from behavior questionnaire and Detailed Surveys including TUE, Comprehensive Health Check, Mental Health and Lifestyle Survey, and Pregnancy and Birth Survey. Its study protocol was published in 2012 [2] . FHMS is funded by the central government [3] and commissioned by the prefectural government to the prefectural-run Fukushima Medical University (FMU) [4].

Screening protocol

TUE consists of the primary examination by thyroid ultrasound screening and the confirmatory examination, if necessary, including more detailed ultrasound examination and urine/blood testing and possible biopsy when needed. The first round of TUE was scheduled to be conducted from October 9, 2011 through March 31, 2014, with each fiscal year from April to the following March covering residents from a set of municipalities grouped according to the air dose level of radiation. 

The second round was scheduled to begin in April 2014, immediately after the first round completed, including residents who were born between April 2, 2012 and April 1, 2013. However, in reality, the primary examination from the first round continued another year through April 30, 2015, concurrent with the second round examination scheduled from April 1, 2014 through March 31, 2015. (FHMS allowed the first timers to participate in the first round even though the second round was going on, as long as they hadn’t received notification for the second round, in order to raise the participation rate of the first round TUE. This effort increased the participation rate by 1.5% to the final participation rate of 81.7%).

The unique diagnostic categories of A1, A2, B and C for TUE were established by the "Diagnostic Criteria Inquiry Subcommittee of Thyroid Examination Advisory Committee," consisting of the following seven organizations: Japan Thyroid Association; Japan Association of Endocrine Surgeons; Japan Association of Thyroid Surgery; The Japan Society of Ultrasonics in Medicine; The Japan Society of Sonographers; The Japanese Society for Pediatric Endocrinology; and Japan Association of Breast and Thyroid Sonology. These diagnostic categories are: 

  • A1: no nodules or cysts found
  • A2: nodules ≦ 5.0 mm or cysts ≦ 20.0 mm
  • B: nodules ≧ 5.1 mm or cysts ≧ 20.1 mm
  • C: requiring immediate secondary examination

(“Cysts” in the TUE are said to be colloid cysts with no malignant potential, as cysts with solid components are classified as “nodules” by the size of the cysts themselves. In other words, a 20.0mm cyst with a solid component would be classified as a 20.0mm nodule and thus placed in the B category).

There was one problem: the lack of baseline data for comparison. Such a large-scale thyroid cancer screening in unexposed children has never been conducted in the world. The FMU officials determined that the screening conducted in the first 3 years after the Fukushima Daiichi nuclear power plant accident be considered baseline [Note] on the premise that the data obtained during this 3-year period would not reflect the effect of radiation exposure since the radiation-induced thyroid cancer only began to appear about 4 years after the Chernobyl accident, establishing the latency of radiation-induced thyroid cancer in children to be about 4 years. Thus the first screening was named “Initial Screening” and later renamed “Preliminary Baseline Screening.” 

  • Note: However, calling the screening conducted after exposure “baseline” does not seem like an appropriate methodology. This presumes any thyroid cancer cases detected in Initial Screening to be due to not radiation effects but screening effect: detection of latent thyroid cancer already present before the accident that would not have been discovered without the screening activity. Can such presumption hold up? In general, radiation-induced cancer seems to refer to cancer whose growth is initiated due to exposure to ionizing radiation as a carcinogen. What if the growth of the pre-existing cancer gets promoted due to radiation exposure? Why would that not be considered radiation effect?

Thyroid ultrasound examination results

As this was the first time such a large-scale thyroid ultrasound screening examination was conducted, each set of the results, released by the Oversight Committee approximately every 3 months beginning on January 25, 2012, caused quite a stir: the public was initially concerned with any ultrasound findings reported, while the officials claimed some of the findings, such as nodules and cysts, were only detected due to high sensitivity of the modern ultrasound equipment and could be physiological and transient. 

The first report [5] officially translated into English, from the Eighth Oversight Committee [6] held on September 11, 2012, shows the rate of A2 at 35-43% and B at 0.5-0.6% for each screening fiscal year (FY). Subsequent reports show a generally increasing tendency for the proportion of A2 from FY 2011 to FY 2013, with the final report [7] of the first round, now called Preliminary Baseline Screening, showing the A2 proportion of 36.4% for FY 2011, 44.6% for FY 2012, and 55.5% for FY 2013, with an overall average of 47.8%. The vast majority (over 98%) of A2 are cysts. Incidentally, the most recent February 2016 second round screening results [8] show the average A2 proportion of 58.5%, slightly higher than the first round. The proportion of B increased from year to year, at 0.5% for FY 2011, 0.7% for FY 2012, and 0.9% for FY 2013, with an overall average of 0.8%. The second round so far shows the B proportion of 0.8 to 0.9%.

Thyroid cancer cases

The first cancer case was reported at the Eighth Oversight Committee meeting held on September 11, 2012. It’s not clearly indicated in the reported results [9] per se, but the minutes of the proceeding (unavailable in English) refer to “one cancer case confirmed after biopsy was conducted in 14 individuals.” Reporting of the biopsy results began, as the confirmatory examination progressed, at the Eleventh Oversight Committee meeting held on June 5, 2013: what was reported included the number of cases suspicious for cancer fine-needle aspiration cytology as well as the number of surgically confirmed cases. Each subsequent reporting of the results revealed an increasing number (14 to 16 more each time) of malignant or suspicious cases, but the number of surgically confirmed cancer cases increased at a slower rate, as surgeries were usually scheduled at the discretion of patients’ life priorities. (Final confirmation of thyroid cancer usually requires pathological examination of the tissue from the resected thyroid gland, and the biopsy results normally only lead to “suspicion” of cancer).

Officials maintained that these findings constituted “screening effect,” that is, widespread screening of asymptomatic individuals often leads to discovery of “latent” cancer that would not have been found if it weren’t for screening. 

As the first round screening wound down, with the primary examination nearly complete and the confirmatory examination progressing further, the second round screening, which began in April 2014, started to show cases suspected or confirmed of cancer. The first thyroid examination report from the second round screening [10] was released at the Seventeenth Oversight Committee meeting held on December 25, 2014 [11], showing 4 cases suspected of cancer. Less than 2 months later, on February 12, 2015, this increased to 8 cases suspected of cancer of which one was surgically confirmed as thyroid cancer [12]. Three months later on May 18, 2015, this nearly doubled to 15 cases suspected of cancer of which 5 were confirmed cancer cases, and yet three months later on August 30, 2015, 10 more were added so there were 25 cases suspected of cancer including 6 cases confirmed as thyroid cancer. November 30, 2015 report revealed 39 cases suspected of cancer, 15 of which have been surgically confirmed as cancer. The most recent data [13] released on February 15, 2016, show 51 suspected cancer cases including 16 surgically confirmed cancer cases.



At the 58th Annual Meeting of Japan Thyroid Association, held November 5-7, 2015, in Fukushima City, Fukushima Prefecture, Dr. Shunichi Yamashita is said to have pointed to screening effect to explain the current increase in thyroid cancer cases. 

However, an important fact needs considered: as seen in the bottom row of the table above, 40 of the 51 cases suspected or confirmed of cancer had either no ultrasound findings (25 cases) or only cysts with no malignant potential (15 cases) in the first round screening. This means, either some ultrasound findings were missed in the first round screening, or new lesions appeared since the first round screening and proved to be cancerous. Fukushima Medical University officials claim there were no missed findings, so these cancers must have grown since the first round screening. This means most of the cancer cases detected during the second round appeared in 2-3 years since the first round screening, contradicting the so-called “latency of four years” that the officials heavily rely on.

The latest tally

The table below shows the most recent results (data as of December 30, 2015) released at the Twenty-First Oversight Committee meeting [14,15],  held on February 15, 2016.



Comparison with annual incidence in Japan

Although it is not appropriate to directly compare between prevalence obtained by screening of general population and incidence based on clinical diagnosis, as a reference the 2010 national incidence [16] estimated in Japan for thyroid cancer in ages 0-19 was 3.3 per million for both sexes, 1.0 per million for male, and 5.6 per million for female [17]. 

Assuming all the suspicious FNAC cases are to be confirmed as cancer, excluding  the single case surgically confirmed to be benign lesions, the first round screening data yields a prevalence of 383 per million (115 cancer cases per 300,478 participants) for both sexes for thyroid cancer in those 0-18 years old at the time of the accident. (However, the estimated incidence significantly increases with age, as shown in the table below, from 1.2 per million for age 10-14 to 11.2 per million for age 15-19, or even 31.1 per million for age 20-24, and about half of the Fukushima cases are over age 18 at diagnosis).

2010 Number of thyroid cancer cases in Japan by age and sex



2010 All (including foreigners) population in Japan by age and sex: “All (including foreigners) population” is used for incidence rate calculation.



2010 Thyroid cancer incidence rate in Japan by age and sex (per million)




Comparison with Chernobyl and other parts of Japan

As the only other major nuclear power plant accident, the Chernobyl accident is often used as a point of reference for many aspects of the Fukushima accident. Official positions as to why Fukushima thyroid cancers, unlike the Chernobyl thyroid cancers, are not considered radiation-induced are roughly summarized in the following 5 points:

  1. Exposure dose is too low (less than 100 mSv above which an increase in cancer occurrence may be statistically shown) in Fukushima.
  2. Unlike Chernobyl where children kept consuming contaminated food, such as milk, internal exposure through consumption of contaminated milk was minimal in Japan due to regulation of food distribution.
  3. In Fukushima, no children under age 5 at exposure have so far been diagnosed with thyroid cancer and latency of the diagnosed cases is too short (therefore the cancer must have already been present at the time of the accident).
  4. Occurrence of ultrasound abnormalities and thyroid cancer in Fukushima Prefecture is comparable to other, “unexposed” areas of Japan.
  5. Genetic analyses of the Fukushima thyroid cancers show a pattern dissimilar to the Chernobyl radiation-induced cancer cases [18].

Point 1: Whereas the Chernobyl exposure doses, often directly measured and swiftly recorded shortly after the Chernobyl accident, might have been significantly higher than the Fukushima exposure doses, the fact is that only 1,083 direct thyroid measurements were conducted in children after the Fukushima accident. Unfortunately, it is an undeniable fact that the reliability of these measurements is questionable due to high background radiation levels. These simple thyroid measurements were intended to be a quick survey, with more detailed testing promised if needed. However, one child from Iwaki City who showed the highest exposure dose of 35 mSv [19] never received any further monitoring: the reason was so as not to “worry and scare” the family and the community. For most, the true exposure dose to radioactive iodine is not known. More detailed diet and behavior history, even at least for those diagnosed with thyroid cancer, might lead to a more accurate dose reconstruction, but  it has not been done.

Furthermore, there are a number of studies showing radiation effects at much lower doses than 100 mSv [20,21,23,24,25].

Regarding point 2, nearly a week had elapsed since the accident by the time the central government established the provisional regulation values for food on March 17, 2011. Meanwhile, raw milk collected in Kawamata Town, Fukushima Prefecture as early as March 16, 2011, showed radioactive iodine levels exceeding the provisional regulation value for milk/milk products of 300 Bq/kg [26]. However, the testing results of the Fukushima raw milk as well as the Ibaraki spinach were not publicized until March 19, 2011 [27]. In the post-earthquake chaos and disruption of food distribution, some might have consumed untested local water, milk, leafy vegetables and other produce which might have been contaminated with high levels of radioactive iodine. Moreover, even when contaminated food might have been avoided, exposure via inhalation might have been unavoidable especially when there was no warning against the approach of the radioactive plume.

As for point 3, in Chernobyl, official stance is that children younger than 5 at exposure began to be diagnosed with thyroid cancer beginning in 1990, the fourth year after the accident. So far in Fukushima, at 4 years after the accident, no cancer case has been seen in children age 5 or younger at exposure. However, TUE is still ongoing for the year 4, with the announcement of the results lagging about 2 months behind the date those results are actually confirmed. No cancer case has been found in children age 5 or younger at exposure in the evacuated municipalities in the 20-30 km zones, but a municipality such as Iwaki City, located in the southern part of Fukushima Prefecture, south of the Fukushima Daiichi NPP, is still undergoing the second round TUE. Iwaki City is a place where unsuspecting residents went about their post-earthquake days, taking care of necessities, lining up outside for water rations, and waiting outside stores for their turns to go inside to purchase needed goods, often with children in tow, totally unaware of the radioactive plume permeating through their city when the wind turned south. Those residents do not know how much radiation they were exposed to from breathing in the contaminated air when the plume came. Lack of post-accident precipitation in Iwaki City, unlike in Iitate Village, means the lack of surface deposition of radioactive substances: the radiation testing of the soil does not reflect the degree of the early exposure doses sustained by residents.

Point 4 refers to the so-called control study [28,29] in Yamanashi, Nagasaki and Aomori Prefectures (a.k.a. the 3-prefecture study)  in which the sample size is much smaller (4,365 vs. 360,000 in Fukushima), and the age distribution and gender proportion are different from the Fukushima study. Although widely (and almost too eagerly) referred to as a control study, it may not really be an appropriate comparison study due to the degree of uncertainty stemming from a large variance from the small sample size: a single case of thyroid cancer diagnosed in the 3-prefecture study makes a point estimate of 229 per 1 million (95% CI: 6 to 1,276 per million).

Genetic analyses mentioned in Point 5 do not constitute a definite proof of radiogenicity and can be influenced by other factors. As a matter of fact, no clear and convenient “fingerprint” exists that can discern radiation effects at this time, although more research is underway [30].

Surgical and pathological features 

Even though TUE is funded by the central government (and administered by the prefectural government), once the participant progresses into the confirmatory examination and needs a closer clinical follow-up, biopsy and/or surgery, the case becomes part of regular medical care under the national health care system. Because biopsy and cancer cases are no longer considered part of TUE, clinical details, such as presence/absence of symptoms, family history, and pathological and molecular genetic findings of thyroid cancer cases are not openly shared for protection of patient privacy. 

The only information reported at quarterly Oversight Committee meetings include age and sex distribution, tumor diameter range, and the types of thyroid cancer (Two types--  papillary thyroid cancer and poorly differentiated thyroid cancer—have been reported so far). During committee proceedings and post-committee press conferences, questions regarding symptoms are often asked by other committee members or journalists. The answer has been consistently, “No symptoms.”

In addition, there have been two reports on surgical and pathological features of thyroid cancer cases operated at FMU. The first was released in November 2014 [31] at the 4th Thyroid Examination Evaluation Subcommittee meeting. The second report was released in August 2015 [32] at the 20th Oversight Committee meeting. Both reports were prepared in response to doubts about over treatment and complaints about lack of clinical data release from the committee members.  

Furthermore, some data have been presented at domestic academic meetings without being released to the prefecture. Abstracts available online are usually in Japanese, but they have been unofficially translated, along with the two reports mentioned above [33,34,35]. 

Pieces of information from different sources are summarized:
As of March 31, 2015, pre-surgical diagnosis revealed that 33 of 96 surgically confirmed thyroid cancer cases had a diameter of 10 mm or smaller. (Surgical treatment of papillary thyroid cancer 10 mm or smaller, called papillary thyroid microcarcinoma or PTMC, is controversial in adults). 8 cases had nodal/distant metastasis or mild extrathyroidal extension. 22 of remaining 25 had proximity to vital organs such as trachea or recurrent laryngeal nerve or cancer cells extending beyond the capsular covering of thyroid gland. In other words, excluding 3 cases which underwent surgeries against recommendations of non-surgical observation, 30 PTMC cases had indications for surgery. Post-surgically, there were 42 PTMC including 14 with mild extrathyroidal extension and 8 with no nodal/distant metastasis or extrathyroidal extension. Overall, 39% had mild extrathyroidal extension and 74% had nodal metastasis.

Below are excerpts from translation of abstracts for presentation at the 27th Annual Congress of the Japan Association of Endocrine Surgeons [36].  The number of cases described differs among them since each study looked at dataset at various points of time:

“(…) there were 84 cases (96.6%) of papillary thyroid cancer amongst 87 surgical cases of pediatric and adolescent thyroid cancer at the end of 2014. They included 3 cases of follicular variants and 4 cases of cribriform-morular type. The solid variant, seen in high frequency after the Chernobyl accident, is classified as poorly differentiated thyroid cancer in the Sixth Edition of Thyroid Cancer Management Guideline.”

“(…) 65 surgical cases of pediatric and adolescent papillary thyroid cancer: 22 males and 43 females; average age 17.4 years; 59 cases of classic subtype, 2 cases of follicular variant, and 4 cases of cribriform-morular type. “

“Surgical methods included total thyroidectomy in 6 cases (8%) and hemithyroidectomy in 73 cases (92%). Lymph node dissection was conducted in all cases, with 82% limited to the central compartment and 18% including the central and lateral compartments. Post-operative pathological diagnosis revealed 17 cases (22%) with tumor diameter ≤ 10 mm, and 44% with extrathyroidal extension, pEx1*, and 75% with lymph node metastasis.”

Although some information can be sought out which provide bits and pieces of information, without having exact and comprehensive details of each cancer case, such as age, sex, municipality of residence at the time of the accident, size and location of tumor, a state of nodal/distant metastasis, and a degree of invasiveness, it is difficult to conduct a further analysis. Lack of sufficient exposure dose information is hailed as one of the main reasons for not being able to conduct a dose-response analysis. In that respect, even a general idea of where the patient was when the radioactive plume came might give a clue to the dose range. 

Release and Analysis of data

FMU and Fukushima Prefecture have not conducted their own epidemiological analysis of the thyroid cancer data. Nor have they released all the available data to make a complete third-party analysis possible. FMU has even prioritized presentations of previously withheld information at academic conferences. Some journalists have repeatedly requested, in vain, the release of information that might offer a clue to any relationship of specific cancer cases with the place of residence as a surrogate for exposure doses. Data released do include the gender and age distributions and the place of residence, without possibility to cross-reference: only the total number of cases is available on the municipality-basis, with no way of knowing the gender and/or age of specific cases. Clinical details of each case are said to be beyond the scope of the Oversight Committee, since the confirmatory examination transitions some cases (biopsy and beyond) from the government-paid screening by the TUE team to the regular medical care by specialists through the national health insurance incurring self-pay costs. At this level, the privacy wall is reinforced, and information from individual cases is not necessarily collected centrally by the prefecture.

In October 2015, the first epidemiological analysis [37] of the publicly available thyroid cancer data (the first round screening data as of December 31, 2014) was published by Tsuda et al. in the online, ahead-of-print edition of Epidemiology, the official, peer-reviewed journal of the International Society for Environmental Epidemiology. The study by Tsuda et al. found a regional variability of the prevalence within Fukushima Prefecture as well as increased incidence rate ratios in most of Fukushima Prefecture compared to the national incidence rate.  Despite the claim by the authors that the study used standard epidemiological methods based on the concept of the discipline of modern epidemiology, it created quite a controversy. There have been criticisms from within and outside Japan [38,39,40,41,42,43,44]. A counterargument by Tsuda et al. has also been published [45].

A group of researchers from the National Cancer Center recently published their analysis [46] and showed the observed/expected ratio of thyroid cancer prevalence to be as much as 30.8. However, they attribute this increase to overdiagnosis.

Jacob et al. (2014) [47] estimated the prevalence of the first round screening and then determined the screening factor for the subsequent screenings. However, a careful consideration of the studies cited by Jacob et al. reveals that data used in estimation was derived from the data obtained 12 to 14 years post-Chernobyl, unlike the first several years post-Fukushima, and involved other factors potentially leading to large uncertainties.

Potential issues

Publicly available TUE data is limited, and the official English translation that is eventually provided may not include the entire data. Additional information might be extracted during the Oversight Committee meeting or the subsequent press conference, but the official minutes, only available in Japanese, do not include the press conference material. Information presented at domestic academic meetings may be available online, but often only in Japanese. All these make it difficult for non-Japanese speakers to obtain thorough information.

Given the fact that the second round has not completed, some say it is too premature to draw any definite conclusion from the data. Ideally, unbiased, collaboratory effort amongst clinicians and researchers to integrate all the available information might lead to a more effective and congruent analytical process that could be useful towards policy making to benefit the public. Such information might include the exposure dose (with a more comprehensive effort to conduct dose reconstruction), the TUE results, and clinical data such as surgical and pathological details. Rather, in reality, various parties are presenting and defending their own claims with little interdisciplinary crossover, reflecting vertical divisions permeating the Japanese society.  

What to think of all this

Radiation epidemiologists and others think that it is premature to determine if the thyroid cancer cases detected in Fukushima children are due to the radiation exposure from the Fukushima Daiichi nuclear power plant accident, as the conventionally accepted latency for childhood thyroid cancer is about 5 years. 

One of UNSCEAR’s conclusions from the 2013 report [48], “No discernible increases in future cancer rates,” is upheld in the 2015 White Paper [49], as presented at the February 9-10, 2016 Public Dialogues held in Fukushima Prefecture [50]. Meanwhile the second round screening is identifying more cancer cases than can be explained by screening effect which should not play a large role due to harvest effect of most latent cancers having been “harvested” in the first round. At the aforementioned Public Dialogues, UNSCEAR officials cited screening effect as an explanation for the thyroid cancer cases. UNSCEAR’s 2015 White Paper only included update information from October 2012 to December 2014, and the second round screening results were not considered.

On January 22, 2016, the International Society for Environmental Epidemiology sent an open letter to the Japanese government [51] expressing their concern about a “12-fold higher risk of developing thyroid cancer among residents of Fukushima” compared to the Japan’s annual incidence, as demonstrated in the study by Tsuda et al. ISEE called for the need to develop scientific studies of health risks from the accident and offered to the government of Japan its expertise as an independent international professional organization of environmental epidemiologists. To date, the Japanese government is yet to acknowledge the ISEE letter [52].

With the report of thyroid cancer cases outside Fukushima Prefecture [53], it is critical for the public health sector to be ready for what might be coming. Assistance from independent bodies of experts would seem wise and desirable.


Yuri Hiranuma, D.O. 
Portland, Oregon, U.S.A.
yurihrnm@gmail.com


References
  1.   http://fmu-global.jp/fukushima-health-management-survey/
  2.   http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3798631/
  3.   http://www.env.go.jp/chemi/rhm/support.html
  4.   http://clearinghouse.main.jp/wp/?p=738
  5.   http://fmu-global.jp/?wpdmdl=37
  6.   http://fmu-global.jp/survey/proceedings-of-the-8th-prefectural-oversight-committee-meeting-for-fukushima-health-management-survey/
  7.   http://fmu-global.jp/?wpdmdl=1222
  8.   http://fmu-global.jp/?wpdmdl=1563
  9.   http://fmu-global.jp/?wpdmdl=37
  10.   http://fmu-global.jp/?wpdmdl=158
  11.   http://fmu-global.jp/survey/proceedings-of-the-17th-prefectural-oversight-committee-meeting-for-fukushima-health-management-survey/
  12.   http://fmu-global.jp/?wpdmdl=170
  13.   http://fmu-global.jp/?wpdmdl=1563
  14.   Ibid.
  15.   http://fukushimavoice-eng2.blogspot.com/2016/02/fukushima-thyroid-examination-february.html
  16.   http://ganjoho.jp/en/professional/statistics/table_download.html
  17.   Ibid.
  18.   http://www.nature.com/articles/srep16976
  19.   http://www.bioone.org/doi/10.1667/RR13351.1
  20.   http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2009418/
  21.   http://www.bmj.com/content/331/7508/77
  22.   http://ehp.niehs.nih.gov/1408548/
  23.   http://www.bmj.com/content/346/bmj.f2360
  24.   http://www.thelancet.com/journals/lanhae/article/PIIS2352-3026%2815%2900094-0/fulltext
  25.   http://www.bmj.com/content/351/bmj.h5359
  26.   http://www.maff.go.jp/j/kanbo/joho/saigai/seisan_kensa/pdf/2011_3g.pdf
  27.   http://www3.nhk.or.jp/news/genpatsu-fukushima/20110319/2010_s_shokuhin_taiou.html
  28.  http://journals.plos.org/plosone/article?id=10.1371/journal.pone.0083220
  29.   http://www.nature.com/articles/srep09046
  30.   http://link.springer.com/article/10.1007/s00259-015-3303-3
  31.   https://www.pref.fukushima.lg.jp/uploaded/attachment/90997.pdf
  32.   https://www.pref.fukushima.lg.jp/uploaded/attachment/129308.pdf
  33.   http://fukushimavoice-eng2.blogspot.com/2014/11/details-of-fukushima-thyroid-cancer.html
  34.   http://fukushimavoice-eng2.blogspot.com/2015/06/2015-update-details-of-fukushima.html
  35.   http://fukushimavoice-eng2.blogspot.com/2015/09/surgical-and-pathological-details-of.html
  36.   http://fukushimavoice-eng2.blogspot.com/2015/08/3-thyroid-cancer-cases-diagnosed-in.html
  37.  http://journals.lww.com/epidem/Fulltext/2016/05000/Thyroid_Cancer_Detection_by_Ultrasound_Among.3.aspx
  38.  http://journals.lww.com/epidem/Citation/publishahead/Re___Thyroid_Cancer_Among_Young_People_in.99055.aspx
  39.  http://journals.lww.com/epidem/Citation/publishahead/Re___Thyroid_Cancer_Among_Young_People_in.99056.aspx
  40.  http://journals.lww.com/epidem/Citation/publishahead/Re___Thyroid_Cancer_Among_Young_People_in.99058.aspx
  41.  http://journals.lww.com/epidem/Citation/publishahead/Re__Thyroid_Cancer_Among_Young_People_in.99064.aspx
  42.  http://journals.lww.com/epidem/Citation/publishahead/Re___Thyroid_Cancer_Among_Young_People_in.99065.aspx
  43.  http://journals.lww.com/epidem/Citation/publishahead/Re___Thyroid_Cancer_Among_Young_People_in.99066.aspx
  44.  http://journals.lww.com/epidem/Citation/publishahead/Re__Thyroid_Cancer_among_Young_People_in.99063.aspx
  45.  http://journals.lww.com/epidem/Citation/publishahead/Response_to_the_Commentary_by_Professor_Davis_and.99060.aspx
  46.   https://jjco.oxfordjournals.org/content/early/2016/01/10/jjco.hyv191.full
  47.   http://link.springer.com/article/10.1007/S00411-013-0508-3#/page-1
  48.   http://www.unscear.org/unscear/en/publications/2013_1.html
  49.   http://www.unscear.org/unscear/en/publications/Fukushima_WP2015.html
  50.   http://www.unis.unvienna.org/unis/en/pressrels/2016/unisma129.html
  51.   http://www.iseepi.org/documents/Fukushimaletter.pdf
  52.   http://www.isee-europe.com/blog/open-discussion-on-isees-letter-to-the-authorities-in-japan-concerning-the-paper-by-tsuda-et-al-2015
  53.   http://fukushimavoice-eng2.blogspot.com/2015/08/3-thyroid-cancer-cases-diagnosed-in.html