Re: “Low-Dose Childhood Radiation Effects to the Thyroid Follow a Linear Dose–Response Trend and Persist Even 45+ Years After Exposure” ( Clin Thyroidol 2017;29:235–236)

Dear Editor:

In her letter (1), Leung comments on the paper by Lubin et al. (2), which presents the findings of a pooled study of nine cohort and case-control studies. The studies were on 252 cases among irradiated individuals and 142 cases among non-irradiated individuals. Lubin et al. reported that for both <0.2 Gy and <0.1 Gy, relative risk increased with thyroid dose (p < 0.01) without significant departure from linearity (p = 0.77 and p = 0.66, respectively) and concluded that their analyses reaffirmed linearity of the dose–response trend. Leung states that the study by Lubin et al. “provides strong supporting evidence that even at radiation doses considered relatively low (as low as <0.1 Gy), there remains a linear dose–response trend of increased thyroid-cancer risks even 45 years after the initial radiation exposure” (1).

Leung's letter and the article by Lubin et al. both have a major shortcoming. They analyzed the dose dependence of thyroid-cancer incidence, which is not clinically significant. It is well known that increased incidence (overdiagnosis) results from screening to detect thyroid nodules. The variability of thyroid-cancer incidence can be very large, depending on the screening effort.

Lubin et al. (2) tried to address this issue by comparing relative risks from earlier time periods to later time periods, implying increased screening at later time periods. They stated that there was no change in the dose–response shape between the periods, indicating overdiagnosis did not likely skew the results. Since there was overdiagnosis of thyroid cancers in earlier periods as well, the consistency of data between early and later time periods does not lead to any conclusion regarding the influence of overdiagnosis on dose–response trends. Incidence is not an appropriate measure of harm to the public from radiation exposure because of the problem of overdiagnosis; thyroid-cancer mortality rates should have been studied.

A recent Korean study by Ahn et al. (3) states, “In 2011, the rate of thyroid-cancer diagnoses was 15 times that observed in 1993. This entire increase can be attributed to the detection of papillary thyroid cancer. Furthermore, despite the dramatic increase in incidence, mortality from thyroid cancer remains stable—a combination that is pathognomonic for overdiagnosis.” A considerable amount of harm has been caused to the Korean public by screening for thyroid cancer due to identification and treatment of indolent thyroid cancers. “Virtually all of the people diagnosed with thyroid cancer are treated: roughly two thirds undergo radical thyroidectomy, and one third undergo subtotal thyroidectomy.” The investigators have asked the health systems around the world to consider their experience with screening as a caution, warning about the business factor. Thyroid-cancer screening and treatment have become big business in South Korea. Over the past decade, hospitals have expanded thyroid clinics, hiring surgeons, and fueling an industry of robot-assisted thyroid surgery.

Cuttler and Pollycove (4) also addressed the effects of screening on the increased thyroid-cancer rates observed after the Chernobyl accident. Considering the issue of “overdiagnosed thyroid cancers,” it is reasonable to suggest that screening and overdiagnosis were higher when the estimated doses were higher. In this light, the findings of Lubin et al. (2) could be severely affected by overdiagnosis.

The exclusive focus on rare cancers such as thyroid cancer presents a completely misleading impression on the effect of radiation on cancer. The study by Franklyn et al. on cancer rates following radioiodine treatment of 7417 hyperthyroid patients showed there was a significant increase in rare thyroid-cancer mortality (5). However, 634 cancer diagnoses were made, compared to an expected number of 761 (standardized incidence ratio = 0.83 [confidence interval (CI) 0.77–0.90]). The relative risk of cancer mortality was also decreased (observed cancer deaths 448, expected 499; standardized mortality ratio = 0.90 [CI 0.82–0.98]). “The decrease in overall cancer incidence and mortality in those treated for hyperthyroidism with radioiodine is reassuring” (5). Paying exclusive attention to thyroid cancers in this study would have given a completely misleading picture on the health impact of the radiation exposures. These patients received a mean total body dose of 54 mGy and a mean thyroid dose of 308 Gy (4).

In summary, Lubin et al. (2) examined the dose–response trend of thyroid-cancer incidence, which is highly susceptible to the screening effort. Their conclusions and claims are deeply flawed. They sustain unwarranted fear of radiation-induced cancer, which seriously impedes very important applications of low doses of radiation in medical diagnostics and treatments.

Finally, an error is apparent in the letter by Leung, Methods section, which states that the Lubin et al. study comprised 252 cases among irradiated individuals with >2.5 million person-years of follow-up and 142 cases among non-irradiated individuals with >1.8 million person-years of follow-up. In table 1 of the article by Lubin et al., it is clear that the person-years of follow-up apply to the 61,155 exposed and the 46,439 non-exposed individuals, not to the 252 and 142 thyroid cancer cases, respectively.