Thyroid Cancer and Benign Nodules After Exposure In Utero to Fallout From Chernobyl

Abstract

Background

Children and adolescents exposed to radioactive iodine-131 (I-131) in fallout from the 1986 Chernobyl nuclear accident appear to be at increased risk of thyroid cancer and benign thyroid nodules. The prenatal period is also considered radiosensitive, and the fetal thyroid can absorb I-131 from the maternal circulation.

Objectives

We aimed to estimate the risk of malignant and benign thyroid nodules in individuals exposed prenatally.

Methods

We studied a cohort of 2582 subjects in Ukraine with estimates of I-131 prenatal thyroid dose (mean = 72.6 mGy), who underwent two standardized thyroid screening examinations. To evaluate the dose-response relationship, we estimated the excess OR (EOR) using logistic regression.

Results

Based on a combined total of eight cases diagnosed at screenings from 2003 to 2006 and 2012 to 2015, we found a markedly elevated, albeit not statistically significant, dose-related risk of thyroid cancer (EOR/Gy = 3.91, 95% CI: –1.49, 65.66). At cycle 2 (n = 1,786), there was a strong and significant association between I-131 thyroid dose and screen-detected large benign nodules (≥10 mm) (EOR/Gy = 4.19, 95% CI: 0.68, 11.62; P = 0.009), but no significant increase in risk for small nodules (<10 mm) (EOR/Gy = 0.34, 95% CI: –0.67, 2.24; P = 0.604).

Conclusions

The dose effect by nodule size, with I-131 risk for large but not small nodules, is similar to that among exposed children and adolescents in Belarus. Based on a small number of cases, there is also a suggestive effect of I-131 dose on thyroid cancer risk.

We studied a cohort in Ukraine exposed prenatally to I-131 in fallout from Chernobyl and found a suggestive increase in thyroid cancer and a significant association with risk of large benign nodules.

Studies of children and adolescents exposed to fallout from the April 1986 accident at the Chernobyl nuclear power plant in Ukraine and Belarus strongly suggest an association between radioactive iodine-131 (I-131) and increased risk of thyroid cancer (1–5), with elevated risk persisting over 30 years after exposure (6). The largest increases are seen among those exposed at the youngest ages, when consumption of I-131-contaminated milk is highest, the small size of the thyroid gland increases the absorbed dose (7), and rapid cell proliferation raises the risk for carcinogenesis. The fetal thyroid gland becomes active around weeks 10 to 12 of gestation, when it begins to accumulate iodine from the maternal circulation via the placental iodine pump (8), and by late gestation, radioiodine concentrations in the fetus may be many times higher than those in the mother (9). Although the in utero period is considered to be especially radiosensitive (10, 11), the risk from prenatal exposure has rarely been studied, and such data are important for establishing guidelines on exposure to pregnant women in medical and occupational settings.

The two prior studies of prenatal exposure to environmental radioiodines and risk of thyroid cancer are only minimally informative owing largely to low statistical power. Subjects exposed in utero to fallout from nuclear tests on the Marshall Islands were very few (N = 4) and their exposure included external radiation and isotopes other than I-131 (12); only 376 in utero–exposed individuals were downwind of the Nevada test site, and no cases of thyroid neoplasia were observed among them in a 30-year follow-up (13). An apparent effect of external radiation exposure on thyroid nodules has, however, been reported among in utero–exposed atomic bomb survivors (14).

Because the data on effects of prenatal exposure are limited and the question is important, in 2002 the US National Cancer Institute, together with collaborators at the Institute of Endocrinology and Metabolism in Kyiv, Ukraine, established a cohort of 2,582 mother-child pairs exposed during pregnancy to I-131 from the Chernobyl accident. From 2003 to 2006, we conducted a cross-sectional thyroid screening examination of the in utero–exposed subjects, who were then between 16 and 19 years of age. There were seven cases of thyroid cancer confirmed cytologically or pathologically, along with 163 ultrasound-detected benign thyroid nodules. Dose-response analyses showed a striking but statistically nonsignificant increased risk of thyroid cancer [excess OR (EOR) per Gy = 11.66; P = 0.12]. No increase in I-131 risk was observed for thyroid nodules (EOR/Gy = –0.31; P = 0.49) (15).

The current study is based on a second screening examination of the Ukraine in utero cohort carried out from 2012 to 2015 when the subjects were in their mid- to late twenties. Here, we present an updated I-131 radiation risk analysis based on the total number of thyroid cancers ascertained in cycles 1 and 2 and a separate analysis of risk for thyroid nodules detected by ultrasonography at the cycle 2 screening.

Methods

Population

The Ukraine in utero cohort consists of mother-child pairs in which the mother was pregnant during the period when I-131 from the Chernobyl nuclear accident was present in the environment (26 April 1986 to 30 June 1986). Details of the study design have been reported previously (15). In brief, an in utero exposed group of subjects (n = 1494) was recruited from three northern oblasts or states (Zhytomir, Chernihiv, and Kyiv) contaminated as a result of the accident (i.e., with cesium-137 (Cs-137) ground deposition greater than 37 kBq/m²). The mothers either had direct thyroid radioactivity measurements themselves made in the months after the accident or lived in the same contaminated settlement as women of child-bearing age with measured thyroid activity that could be used to estimate doses to the unmeasured mothers. A comparison group of subjects (n = 1088) was comprised of offspring born to women pregnant during the same period of time but residing in relatively uncontaminated settlements (Cs-137 ground deposition less than or equal to 37 kBq/m²), principally in raions or counties of the same three oblasts. Standardized procedures for tracing and recruitment resulted in participation rates for the initial thyroid screening (cycle 1) in excess of 80% among eligible offspring from both exposure groups. In the second screening (cycle 2) a decade later, the overall participation rate was 70% (1796/2582). Ten subjects with a history of thyroid surgery prior to the second thyroid screening were excluded from the analysis of thyroid nodule prevalence. The final study sample included 1786 subjects.

Screening examination

The thyroid screenings of the in utero–exposed subjects carried out initially from 2003 to 2006 and again from 2012 to 2015 used standardized procedures developed for the Ukrainian American screening study of 13,243 individuals <18 years of age at the time of the accident, known as UkrAm (16). The screening protocol includes thyroid palpation and ultrasonographic examination by an ultrasonographer and independent clinical examination and palpation by an endocrinologist. Presence of nodules, their size, echostructure, and pattern of echogenicity were recorded. Subjects with nodules 10 mm or greater in the largest dimension, as well as any nodule 5 to 10 mm in size with ultrasound characteristics suspicious for malignancy, were referred for fine-needle aspiration (FNA) biopsy; subjects with FNA findings diagnostic or suspicious for thyroid carcinoma and neoplasia were referred for surgery. Structured questionnaires, administered to the subjects or their mothers, were used to gather information on thyroid disease, x-ray exposure during pregnancy, and items relevant for dose estimation, including residential history, consumption of contaminated foods, and iodine prophylaxis from May to June 1986.

The screening studies were reviewed and approved by the institutional review boards at the US National Cancer Institute and the local collaborating institution in Ukraine, and all the subjects and their mothers signed an informed consent.

Outcome definitions

The cycle 2 results are based on all screening examinations carried out between 2012 and 31 December 2015, as well as follow-up thyroid surgeries through 31 December 2016. In the analyses, we defined several outcomes of interest: (1) thyroid cancer identified and pathologically confirmed at either the first or second thyroid screening and (2) thyroid nodules identified by ultrasound examination at the second screening and further categorized into neoplastic, suspicious, or benign based on a combination of ultrasound, FNA, and histological findings (17).

The present analysis of thyroid cancer is based on eight confirmed cases (six at cycle 1 and two at cycle 2). One case included in our previous report (15) with an FNA conclusion “suspicious for follicular neoplasia” and not operated until the end of 2017 remains unconfirmed and has now been excluded.

Dosimetry

Individual in utero I-131 thyroid doses were calculated for members of the cohort by first estimating the variation with time of I-131 activity in the thyroid of the mother and then, using a biokinetic ecologic model, estimating dose to the embryo/fetus (18).

For mothers with direct thyroid measurements (28% of the total), the thyroid dose due to I-131 intake was derived from the measurement itself. For mothers without direct measurements from contaminated settlements where thyroid I-131 activities were measured in other adult women, the thyroid dose was calibrated by measurements done at that settlement. For mothers without direct thyroid measurements from uncontaminated settlements, the thyroid dose was inferred from the relationship observed between Cs-137 deposition density at that settlement and the thyroid dose in settlements with direct thyroid measurements. The estimate for the mother then served as input to estimating dose to the embryo/fetus based on a bidirectional model developed for the International Commission on Radiation Protection (19). The model accounts for transfer of iodine between the maternal and fetal pools and retention of iodide in the placenta, and predicts a continuous increase in dose with increasing gestational age such that doses are minimal early in gestation when the fetal thyroid is not yet fully active and maximal in the third trimester (20).

Depending on the date of birth, some cohort members born between May and June 1986 might have had exposure to I-131 postnatally. For such subjects, I-131 thyroid dose from postnatal exposure was estimated as described elsewhere (18) and their cumulative thyroid dose was calculated as a sum of fetal and postnatal I-131 doses.

Statistical analysis

To evaluate the association between I-131 thyroid dose and prevalence of thyroid cancer or benign thyroid nodules, we estimated the excess OR (EOR = OR-1) and computed 95% CIs using logistic regression (21) in Epicure (22). The model has the following general form:

$$Odds\left(x,d\right)=p\left(x,d\right)/\left[1-p\left(x,d\right)\right]=r\left(x\right)\left[1+ f\left(d\right)\right]$$

where x is a vector of covariates (e.g., sex, age), d is the cumulative I-131 thyroid dose, $r\left(x\right)$ is a function of background parameters, and $f\left(d\right)$ is a function describing the effect of I-131. Under the main dose-response model, EOR linearly increases with dose, so $f\left(d\right)=\alpha d$ and is expressed as EOR per Gy. To evaluate departure from linearity, we allowed for modification of linear EOR by the exponential term, so that$f\left(d\right)=\alpha d\exp [\beta d]$. Where the numbers permitted, for exploratory purposes we also considered a categorical dose response (with cutpoints of 0.020, 0.050, 0.100, 0.200, 0.400, and 0.800 mGy). To assess the influence of postnatal dose, we repeated analyses excluding 57 individuals with postnatal doses of I-131 ≥30 mGy.

The model parameters and nested models were evaluated using the likelihood ratio test (23). All statistical tests were two-sided with a specified type I error of 0.05.

Results

Among 1786 subjects with internal I-131 thyroid exposure in utero who were examined in cycle 2, 55% were females (Table 1). The mean age at the time of examination was 27 years (range: 25 to 30 years). Sixty-one percent were born to mothers residing in contaminated territories (Cs-137 >37 kBq/m²). The mean cumulative I-131 thyroid dose increased with trimester of exposure from 2.1 mGy in the first semester, 73.4 mGy in the second semester, and 125.8 mGy in the third trimester, while the overall mean I-131 dose was 72.6 mGy (range: 0 to 2268 mGy).

Table 1.

Selected Characteristics of Individuals From the Ukraine In Utero Cohort: Second Thyroid Screening, 2012–2015

Characteristic	n (%) or Mean (SD)
Sex
Female	987 (55)
Male	799 (45)
Age, ya	27 (1)
Mother’s residence
Contaminated territory	1094 (61)
Comparison territory	692 (39)
I-131 thyroid dose, mGyb	72.6 (189.3)
First trimester	2.1 (8.5)
Second trimester	73.4 (157.0)
Third trimester	125.8 (259.6)

^aAt the time of second screening examination.

^bCumulative dose at initial screening including prenatal and, for some individuals exposed during the third trimester, postnatal I-131 doses.

Thyroid cancer, cycles 1 and 2

The two screenings of the in utero–exposed individuals identified eight cases of thyroid cancer – six at cycle 1 and two additional cases at cycle 2. Both new cases were diagnosed in females from relatively uncontaminated areas. Seven of the eight were papillary thyroid carcinomas and one was a follicular carcinoma. The estimated I-131 risk based on the eight cases remained elevated but not statistically significant: EOR/Gy = 3.91 (95% CI: <–1.49, 65.66; P = 0.34).

Thyroid nodules, cycle 2

At the cycle 2 ultrasound examination, at least one thyroid nodule was identified in 241 of 1786 in utero–exposed cohort members (13.5%). Twenty nodular cases were neoplastic or suspicious (8.3%) and 221 were benign (91.7%). Among the latter group, 178 cases (80.5%) were small nodules (<10 mm) and 43 (19.5%) were large nodules (≥10 mm).

In categorical analysis of I-131 risk (Table 2), the EORs for each case group increased with dose up to 0.400 to 0.799 Gy and decreased thereafter. However, we failed to reject a hypothesis of homogeneity of dose-category–specific EORs.

Table 2.

EORs and 95% CIs for Thyroid Nodules Across I-131 Thyroid Dose Categories in the Ukraine In Utero Cohort: Second Thyroid Screening, 2012–2015

Nodule Grouping	I-131 Thyroid Dose, Gy
	<0.020	0.020–0.049	0.050–0.099	0.100–0.199	0.200–0.399	0.400–0.799	0.800–2.268	P Hetera
	0.005	0.03	0.072	0.146	0.269	0.555	1.28
No nodule, n	893	263	148	112	71	36	22
Any nodule, n	124	37	27	22	15	12	4
EOR	Referent	0.02b	0.33	0.42	0.79	1.37	0.30	0.130
95% CI		–0.32, 0.51	–0.17, 1.08	–0.16, 1.30	–0.05, 2.18	0.14, 3.62	–0.63, 2.53
Benign nodule, n	113	36	25	20	14	10	3
EOR	Referent	0.09	0.35	0.41	0.80	1.18	0.08	0.234
95% CI		–0.28, 0.62	–0.17, 1.14	–0.18, 1.32	–0.06, 2.25	–0.01, 3.42	–0.75, 2.22
<10 mm, n	94	28	22	17	10	5	2
EOR	Referent	0.02	0.43	0.41	0.54	0.34	–0.13	0.653
95% CI		–0.36, 0.58	–0.15, 1.32	–0.21, 1.41	–0.28, 2.00	–0.55, 2.24	<–1.00, 2.05
≥10 mm, n	19	8	3	3	4	5	1
EOR	Referent	1.9	0.95	1.49	5.27	12.18	3.16	0.067
95% CI		0.18, 5.51	–0.55, 4.86	–0.42, 6.50	0.76, 16.57	3.12, 34.84	<–1.30, 21.05

^aSix degrees of freedom tests of heterogeneity of EORs.

^bAll estimates are adjusted for sex.

Based on a simple linear model, the EOR per Gy of cumulative I-131 dose for cases with any nodule was 1.53 (95% CI: 0.22, 3.59; P = 0.015) (Table 3). The corresponding estimate for cases with benign nodules was lower and not statistically significant (EOR/Gy = 1.19, 95% CI: –0.09, 2.16; P = 0.075). However, there was a highly significant dose-related increase in I-131 risk for the subgroup of large benign nodules (EOR/Gy = 4.19, 95% CI: 0.68, 11.62; P = 0.009), but no significant increase for small benign nodules (EOR/Gy = 0.34, 95% CI: –0.67, 2.24; P = 0.604). The difference in dose-response slopes for large and small benign nodules was not statistically significant (P = 0.063).

Table 3.

EORs and 95% CIs for I-131 Thyroid Dose by Nodule Grouping in the Ukraine In Utero Cohort: Second Thyroid Screening, 2012–2015

	All Subjects						Subjects With Postnatal I-131 Thyroid Dose <0.30 mGya
Nodule Grouping				Dose Term						Dose Term
	n	EORb per Gy	P Linearc	Lineard	Exponentiale	P Expf	n	EOR per Gy	P Linear	Linear	Exponential	P Exp
Any nodule	241	1.53	0.015	4.92	–1.55	0.038	229	2.01	0.004	3.38	–0.77	0.296
		0.22, 3.59g		1.13, 12.36	–3.98, –0.08			0.49, 4.41		0.64, 9.16	–2.90, 0.67
Benign nodule	221	1.19	0.075	5.48	–1.98	0.021	211	1.71	0.023	3.98	–1.28	0.141
		–0.09, 3.25		1.08, 14.20	–5.12, –0.30			0.18, 4.15		0.61, 11.24	–4.32, 0.41
<10 mm	178	0.34	0.604	6.20	–3.16	0.039	171	0.79	0.307	4.33	–2.34	0.131
		–0.67, 2.24		–28.05, 28.17	–22.86, –0.22			–0.52, 3.15		–12.86, 34.59	NE, 1.07
≥10 mm	43	4.19	0.009	7.34	–0.83	0.324	40	5.04	0.005	5.82	–0.24	0.781
		0.68, 11.62		<–2.44, 25.54	–3.12, 0.79			0.98, 13.73		0.75, 20.72	–2.44, 1.37

Abbreviations: exp, exponential; NE, not estimable.

^aAnalyses are based on cumulative I-131 thyroid dose (prenatal plus any postnatal component) for all subjects and then restricted to subjects with postnatal I-131 dose <30 mGy.

^bEOR per Gy is estimated from simple linear model and adjusted for sex.

^cOne degrees of freedom test for linear trend.

^dLinear dose term from linear-exponential model adjusted for sex.

^eExponential dose term from linear exponential model adjusted for sex.

^fOne degrees of freedom test for exponential departure from linearity.

^g95% CI.

Further evaluation of the shape of the dose response revealed evidence of statistically significant departure from linearity in I-131 risk for all nodules, all benign nodules, and small benign nodules (P < 0.05), with risk reducing at high doses; large benign nodules were the exception (P = 0.324) (Table 3). Because high-dose individuals were mainly exposed in the third trimester and some had postnatal exposure to I-131, we conducted sensitivity analyses excluding individuals considered to have nonnegligible postnatal thyroid doses (≥30 mGy) (Table 3). The EOR/Gy for each case group in the restricted dose range became somewhat higher, but the pattern of I-131 risk remained the same: 2.01 (95% CI: 0.49, 4.41; P = 0.004) for any nodule, 1.71 (95% CI: 0.18, 4.15; P = 0.023) for benign nodules; 5.04 (95% CI: 0.98, 13.74; P = 0.005) for large benign nodules, and 0.79 (95% CI: –0.52, 3.15, P = 0.307) for small benign nodules. There was no evidence of significant departure from linearity in the restricted dose range for any case group (Table 3) or significant difference in dose response slopes for large and small benign nodules (P = 0.086).

Discussion

In a relatively large, well-established cohort exposed prenatally to I-131 in Ukraine, with individual estimates of thyroid dose and standardized thyroid screening examinations, we found a markedly elevated, but not statistically significant, dose-related risk of thyroid cancer. Cases were ascertained over two screening cycles, the first from 2003 to 2006 and the second from 2012 to 2015. Analysis of cycle 2 data also showed a strong and significant association between I-131 thyroid dose and screen-detected, large benign thyroid nodules (≥10 mm).

Thyroid cancer

There was no statistically significant excess I-131 risk of thyroid cancer among the in utero–exposed cohort in Ukraine, despite consistently large central estimates: EORs of 11.66 at cycle 1 and 3.91 for cycles 1 and 2 combined. This is largely due to the small number of cases, which limited our power to detect significant effects. All but two of the cancers arose in individuals exposed after the first trimester, when the thyroid gland is able to accumulate iodine and doses are highest. We view the strong but nonsignificant elevations over the two screening cycles as suggestive of a raised risk for thyroid cancer among the in utero exposed.

Our findings for in utero exposure and thyroid cancer can be compared with reports on the prenatally exposed Japanese atomic bomb survivors. Among 319 who were evaluated clinically in their mid-50s (14), five papillary thyroid cancers were found, with age at diagnosis from 34 to 56 and all occurring in females. Although the exposure in Japan was to acute, external radiation at high dose rates rather than protracted internal exposure to I-131, as in our study, there is also a female excess and a long latency. Given the small number of cases, no dose-response analysis was conducted.

A report based on a larger group of Japanese in utero–exposed survivors (N = 2452) compared risks of all adult-onset solid cancers among those exposed prenatally with those exposed before the age of six (24); there were too few thyroid cancers to analyze separately. Both the in utero and early childhood groups showed statistically significant dose-related increases in solid cancer incidence. Prior to their early 20s, the exposed in utero subjects had higher excess RRs per Sievert (ERRs/Sv) for solid cancer than those exposed in early childhood. However, later in life, the ERRs/Sv were lower following in utero exposure than for exposure at young ages.

In our study, we found that the central risk estimates for those exposed to I-131 in utero were statistically comparable to the ERR/Gy of 3.24 (95% CI: not estimable, 539) estimated based on 13 cancers among postnatally exposed children age 1 to 5 in the main UkrAm cohort (15). However, both the prenatal and postnatal I-131 risk estimates have large CIs due to the small number of cases as well as the uncertainty in the estimated doses. Although there is considerable interest in the relative radiosensitivity of the fetal vs young thyroid (25), as yet there is not sufficient epidemiologic evidence to establish a clear difference between the prenatal and early postnatal radiation risks in humans.

Together with collaborators at the Republican Research Center for Radiation Medicine & Human Ecology in Gomel, Belarus, the National Cancer Institute has established a parallel cohort of 2954 individuals exposed in utero to I-131 from Chernobyl accident fallout and who are now receiving standardized thyroid screening examinations, as in Ukraine. Eventual pooling of the Ukraine and Belarus in utero cohorts will increase statistical power of the thyroid cancer analyses and generate more accurate risk estimates for prenatal exposure to I-131. There is also potential for a case-control study among the entire Ukrainian population exposed in utero to fallout from Chernobyl (26).

Benign thyroid nodules

Thyroid nodules—abnormal growths of thyroid tissue—are both very common and usually (>90%) benign, with prevalence higher in females and increasing with age (27, 28). Even nodules that are benign require evaluation and management.

Our finding in the Ukraine in utero cohort of a significant dose-response association with large but not small benign thyroid nodules is consistent with studies in groups exposed to internal or external radiation as children or adolescents. Among 2668 Japanese atomic bomb survivors less than 10 years old at exposure and with known dose (29), a significant association was found between external gamma radiation and large thyroid nodules (≥10 mm), with risk greater for those younger at exposure. The estimates in the Japanese cohort were higher for malignant nodules (EOR/Gy = 4.40, 95% CI: 1.75, 9.97), but were also significantly elevated for benign nodules (EOR/Gy = 2.07, 95% CI: 1.16, 3.39). No association was observed between radiation dose and small (<10 mm) thyroid nodules. An updated analysis of 2376 individuals living near the Semipalatinsk test site in Kazakhstan before age 20, which accounted for dosimetric uncertainties (30), found elevated risks for thyroid nodules with exposure to internal as well as external radiation (ERRs/Gy of 3.59 and 1.47, respectively). A prospective study of 2496 school children exposed to radioactive iodine downwind of the Nevada test site observed a significantly increased risk of thyroid neoplasms and thyroid nodules 30 years after exposure (31), although no cases of thyroid neoplasia were seen in the 376 “downwinders” who were exposed in utero (13).

In the Ukraine in utero–exposed cohort, at cycle 1 no association with dose was observed for thyroid nodules detected when the subjects were in their teens. At cycle 2, however, when the cohort was evaluated in their mid- to late 20s, we did find a statistically significant radiation-related risk for thyroid nodules as a group (EOR/Gy = 1.53, P = 0.015) that was driven by the considerable excess for large benign nodules (EOR/Gy = 4.19, P = 0.009). No association with I-131 dose was observed for small benign nodules (EOR/Gy = 0.34, P = 0.604). A similar pattern by nodule size was found in analyses focused on fetal thyroid dose (i.e., excluding individuals with postnatal dose ≥30 mGy).

Thyroid nodules have also been studied in the 11,970 members of the parallel Belarus American cohort exposed to I-131 from the Chernobyl accident in childhood or adolescence and screened for the first time from 1996 to 2001 (17). Among Belarus American subjects, there was also a significant dose response for all thyroid nodules (EOR/Gy = 0.70 for 5 year olds at the time of the accident, 95% CI: 0.33, 1.18), with risks higher for neoplastic nodules (regardless of size) and for those younger at exposure. For nonneoplastic nodules, there was significantly increased I-131 risk for large (EOR/Gy = 1.55, 95% CI: 0.36, 5.46) but not small thyroid nodules (EOR/Gy = 0.02, 95% CI: <–0.02, 0.47).

Among ∼300 in utero–exposed atomic bomb survivors in Japan who were examined in their mid-50s, elevated radiation risk was observed for large solid nodules (OR/Gy = 2.78, 95% CI: 0.50, 11.80) (14); risk for small nodules was not evaluated. Although not statistically significant, the estimate for large solid nodules was similar to the estimate reported for those exposed in childhood (OR/Gy = 2.65, 95% CI: 1.96, 3.65), and it was suggested it would be “prudent” to assume that a dose-related risk for large solid nodules also exists for those exposed in utero.

The relationship with large but not small thyroid nodules that is quite consistently observed with both external and internal radiation is noteworthy. Whether irradiated large thyroid nodules are prone to progress to malignancy is a concern, particularly to clinicians and those tasked with radiation protection. In a follow-up of 31 atomic bomb survivors with large solid thyroid nodules initially diagnosed as “benign” (32), the hazard ratio (HR) for cancer development was found to be substantially and significantly elevated (HR = 40.2, 95% CI: 9.4, 173.0), while the HR for those with cysts was somewhat elevated but not significantly so (HR = 2.7, 95% CI: 0.3, 22.2). RET/PTC rearrangements and RAS mutations, seen frequently in thyroid cancers, are also sometimes found in benign nodules (33). However, a study of genomic alterations in a series of unirradiated individuals with benign thyroid nodules and concurrent papillary thyroid cancer found that the benign nodules had a different molecular signature and gene expression pattern (34), and this was interpreted to mean that malignant and benign nodules develop independently. Evidence on the natural history of thyroid lesions in irradiated individuals is scant.

Our study of thyroid nodules is one of a very few to look at effects on the thyroid gland from in utero exposure to I-131. Strengths of the study include the careful estimates of both prenatal and postnatal I-131 exposure, which allowed us to carry out detailed dose-response analyses. All subjects, irrespective of dose, went through a thorough ultrasound examination for a variety of thyroid diseases by examiners blinded to dose, thus largely eliminating the potential for surveillance bias. Interviews were carried out with both the cohort members and their mothers to gather relevant information.

There are also some limitations to the study. Although we do have individual estimates of I-131 fetal thyroid dose, these have large uncertainties due to the lack of direct radioactivity measurements on all mothers and the extrapolation from maternal to fetal exposure. Also, the number of thyroid cancer cases was small, and this limited our statistical power to detect moderate to large elevations in risk as significant. The screening cycles were cross-sectional and carried out years after the accident in 1986 (cycle 1 was 12 to 14 years later and cycle 2 close to 30 years after the accident). This means that we were unable to evaluate associations with small thyroid nodules close in time to radiation exposure or to follow them to examine nodule growth and progression. Nonetheless, the findings from a second screening of the Ukraine in utero cohort provide useful data on the prevalence of benign and malignant thyroid nodules following prenatal exposure to I-131.

Conclusions

In summary, screening examinations in a cohort exposed in utero to radioactive iodine in fallout from the Chernobyl nuclear accident in Ukraine have revealed, decades later, a marked but nonsignificantly increased risk of thyroid cancer and a strong, significant dose response for large (≥10 mm) benign thyroid nodules. As pregnant women may be exposed to I-131 in medical, occupational, or environmental settings, the findings from our cross-sectional study are worthy of note, but further research with larger populations and prospective follow-up is needed.

Glossary

Abbreviations:

Cs-137

cesium-137

EOR

excess OR

ERR

excess RR

FNA

fine-needle aspiration

HR

hazard ratio

I-131

iodine-131