Hypothyroidism after radiation exposure: brief narrative review

Abstract

The thyroid gland is among the organs at the greatest risk of cancer from ionizing radiation. Epidemiological evidence from survivors of radiation therapy, atomic bombing, and the Chernobyl reactor accident, clearly shows that radiation exposure in childhood can cause thyroid cancer and benign thyroid nodules. Radiation exposure also may induce hypothyroidism and autoimmune reactions against the thyroid, but these effects are less well-documented. The literature includes only a few, methodologically weak animal studies regarding genetic/molecular mechanisms underlying hypothyroidism and thyroid autoimmunity after radiation exposure. Rather, evidence about radiation-induced hypothyroidism and thyroid autoimmunity derives mainly from follow-up studies in patients treated with external beam radiotherapy (EBRT) or iodine-131, and from epidemiological studies in the atomic bombing or nuclear accident survivors. Historically, hypothyroidism after external irradiation of the thyroid in adulthood was considered not to develop below a 10-20 Gy dose threshold. Newer data suggest a 10 Gy threshold after EBRT. By contrast, data from patients after iodine-131 "internal radiation therapy" of Graves´ disease indicate that hypothyroidism rarely occurs below thyroid doses of 50 Gy. Studies in children affected by the Chernobyl accident indicate that the dose threshold for hypothyroidism may be considerably lower, 3-5 Gy, aligning with observations in A-bomb survivors exposed as children. The reasons for these dose differences in radiosensitivity are not fully understood. Other important questions about the development of hypothyroidism after radiation exposure e.g., in utero, about the interaction between autoimmunity and hypofunction, and about the different effects of internal and external irradiation still must be answered.

Keywords: Autoimmune thyroiditis; Diagnostic medical radiation exposure; Hypothyroidism; Other radiation exposure (atomic bombing/nuclear accidents); Therapeutic medical radiation exposure (EBRT/ RAI); Thyroid.

Fig. 1

Dose–response rate curves (solid line: mean, dotted lines: 95% confidence interval) for hypothyroidism with increasing thyroid doses after RAI in patients with Graves’ disease (authors’ data from a German multicenter study, reconstructed after Peters et al. 1995)

Fig. 2

Mean TSH and mean free thyroid hormone levels from pre-radiotherapy to 48 months post-EBRT in patients with nasopharyngeal cancer (N = 68). The error bars represent the standard deviation (SD) of the hormone levels at the given measurement time. 0m pre-radiotherapy, FT3 free triiodothyronine, FT4 free thyroxine, m month, TSH thyroid-stimulating hormone (Lin et al. , reproduced with permission under the PLOSone Creative Common Attributive Licence)

Fig. 3

Dose–response analysis (blue, red, green, and yellow broken lines) and the overall pooled estimate (solid black line) of 4 studies of the incidence of hypothyroidism in patients receiving EBRT in the neck region for cancer, based on meta-analysis of the steepness of the curve and of the dose to the thyroid resulting in a 50% rate of hypothyroidism in the pooled estimate of 4 studies (Vogelius et al. , with permission)

Fig. 4

Association between prevalence of hypothyroidism and estimated radiation dose to the thyroid: Ukrainian–American cohort study of thyroid cancer and other thyroid diseases after the Chernobyl accident, 1998–2000. The dose–response line was adjusted to pass through the lowest thyroid dose category. Data points represent the thryoid dose category-specific odds ratios (ORs) with 95% CIs. The line represents fitted ORs based on a linear EOR model (Ostroumova et al. , with permission)

Fig. 5

Dose–response association between prevalence of hypothyroidism and estimated thyroid from I-131 in a cohort study of thyroid cancer and other thyroid diseases after the Chernobyl accident in Belarus, 1996–2003. The dose–response line was adjusted to pass through the lowest dose category. The data points represent dose category-specific ORs with 95% CIs (whiskers). Curves represent fitted ORs based on linear (dotted line) and linear-quadratic (solid line) EOR models (Ostroumova et al. , with permssion)