Family History of Thyroid Cancer and the Risk of Differentiated Thyroid Cancer in French Polynesia

Abstract

Background:

Differentiated thyroid carcinoma is considered to be the nonhereditary cancer for which familial inheritance is the highest. To date, no familial aggregation analysis of this cancer has been performed in Maohi populations, which exhibit a very high incidence rate. Therefore, we evaluate the risk of differentiated thyroid cancer associated with a family history of thyroid cancer in natives of French Polynesia.

Methods:

We investigated thyroid cancer incidence in the first-degree relatives of 225 cases of differentiated thyroid carcinomas diagnosed between 1979 and 2004 in patients born in French Polynesia, and 368 randomly selected population controls matched for sex and age, born and residing in French Polynesia. All but five thyroid cancers declared among relatives were validated.

Results:

Twenty-four cases declared a family history of thyroid cancer, when compared with 11 controls. Individuals with an affected first-degree relative had a 4.5-fold (95% confidence interval [CI], 1.9–10.6) increased risk of differentiated thyroid cancer. This odds ratio (OR) was not significantly higher when a male first-degree relative was affected (OR, 10.0; 95% CI, 1.3–74.8) compared with a female (OR, 4.0; 95% CI, 1.5–10.3) and was not different for patients who had a nonaggressive thyroid microcarcinoma (OR, 3.5; 95% CI, 0.6–16.4) than those who had a larger cancer (OR, 6.0; 95% CI, 1.8–20.5). This OR was borderline significantly (p, 0.07) higher in Maohis (OR, 11.0; 95% CI, 2.4–48.8) than in individuals of mixed origin (OR, 2.1; 95% CI, 0.8–5.9).

Conclusion:

Our study shows that the familial inheritance of differentiated thyroid cancer is particularly high in Maohi populations.

Introduction

S ubstantial variations in the incidence of thyroid cancer worldwide strongly suggest that environmental factors are implicated in the etiology of this cancer. The highest thyroid cancer incidence rates are observed in the populations of the Pacific, attaining 17/100,000 among French Polynesian women between 1985 and 1995, which is twice that observed in the Hawaiians of Hawaii and Maoris of New Zealand (1), and threefold higher than that observed among immigrants living in French Polynesia (2).

Ionizing radiation is the main identified risk factor for differentiated papillary thyroid carcinoma (3 –7). In an important study based on the Swedish Family-Cancer Database, thyroid cancer had the highest familial risk among all studied cancer sites, with a familial hazard ratio (FHR) of 10.7 (95% confidence interval [CI], 6.9–16.6) (8,9). Case–control studies on association between a family history of thyroid cancer and a risk of thyroid cancer were based on small numbers of affected relatives and therefore did not always restrict their analysis to first-degree relatives (10 –13). Only one case–control study observed a significant association between having a family history of thyroid cancer and an increased risk of thyroid cancer (odds ratio [OR], 6.1; 95% CI, 1.8–21.3) (10).

To explain the high incidence of thyroid cancer in French Polynesia, we performed a case–control study to investigate the potential role of French atmospheric nuclear testing, which took place between 1966 and 1974 at Mururoa (14), and to study other risk factors for differentiated thyroid cancer risk among the natives in this Pacific population. This study focuses on the role of family history of thyroid cancer among first-degree relatives.

Materials and Methods

The study population and methods have previously been reported (15,16). The study was approved by the French Polynesian Ethics Committee. Written informed consent was obtained from all participants so that we could contact their physician.

Case selection

All patients diagnosed with a differentiated thyroid cancer before the age of 56, born and living in French Polynesia, were eligible. This cutoff for age was chosen to include mainly the subjects who were less than 15 years old at the time of the atmospheric nuclear testing (between 1966 and 1974). The cases were identified from the cancer registry of French Polynesia and medical insurance files and by the four endocrinologists in Tahiti. Histological information was obtained from the two histopathology laboratories in Tahiti and from the medical files of endocrinologists.

Of the 251 eligible differentiated thyroid cancer cases, 26 (10%) were not interviewed because they had died (n = 14), could not be located (n = 6), or refused to participate (n = 5), or the person was too ill to be interviewed (n = 1). Finally, the study population consisted of 225 cases.

Matching process

For each eligible case, two potential controls, matched for the date of birth (±6 months) and sex, were randomly selected from the registry of births, which registers all births in French Polynesia. A SAS random number generator (SAS Institute, Cary, NC) was used for random selection. At the end of the interviews, a few cases had no interviewed controls. Each of those cases was matched with a control initially selected for another case of the same sex and with the same date of birth. Matching on the date of birth was then extended if necessary. Finally, of the 225 interviewed cases, 82 were matched with 1 control (36%) and 143 with 2 controls (64%).

Contacting controls and conducting interviews

Of the 453 randomly selected controls, 85 (19%) were not interviewed because the subjects had died (n = 9), could not be located (n = 29), refused to participate (n = 29), were too ill to be interviewed (n = 2), or had left French Polynesia (n = 16). A total of 368 controls were included.

Data collection

The addresses of cases and controls were obtained from the territorial medical insurance plan. Interviews were conducted face-to-face by trained Polynesian interviewers and medical staff using a structured questionnaire. Data collection included their ethnicity, education, smoking habits, family history of thyroid cancer, gynecological and reproductive factors for women, medical X-ray exposure, weight at age 18, 30, 40, 50 years and at interview, and height at interview. For relatives with thyroid cancer, participants had to report their kinship (father, mother, brother, sister, children), first and last names, year of birth, age at diagnosis of thyroid cancer, type and location of treatment, year of death, and name of attending physician.

Validation of diagnoses of thyroid cancers was done with the endocrinologists or pathologists of Tahiti by ascertaining diagnoses and verifying the histological type, size of the cancer, and the dates of birth and of the diagnosis for each relative with a thyroid cancer. This verification led us to exclude two lesions, a goiter and lung cancer, which were erroneously declared as thyroid cancers.

Among cases, 25 declared having at least one thyroid cancer among their first-degree relatives (see Table 2). The diagnosis was validated for all but five relatives.

Among controls, 11 declared having a thyroid cancer among their first-degree relatives. The diagnosis was validated among all but three of them.

Statistical analysis

Analyses were conducted using a generalized linear model with nonconditional logistic regression (17). This model was applied because different cases included in our study belonged to the same families. To take into account this cluster effect, a family variable was attributed to all cases and controls, which had the same value for all cases belonging to the same family. The correlation matrix chosen for the model was identical supposing there was no correlation in clusters. We applied the GENMOD procedure of the SAS software, version 9.1 (SAS Institute). The model was adjusted for age and sex to take into account the matching variables.

We did not use an aggregation index to correct the observed number of thyroid cancer cases in each family by the expected one (18,19). This was not done for two reasons. First, this method requires the age at onset of thyroid cancer in any familial case, and this value was missing for 16% of the cases in our study. Second, because of the low incidence of thyroid cancer, and the restriction of the analysis to first-degree relatives, the expected number of thyroid cancer cases in each case or control family was much lower than one, meaning that any observed thyroid cancer case in a family would have been in excess. Nevertheless, to take into account the family size, we adjusted our models on the number of siblings as a surrogate for family structure.

Similarly, because of missing data for age and the low incidence of thyroid cancer, we chose the case–control approach instead of comparing the risk of disease among case relatives versus that among control relatives (20). This second approach, referred to as the cohort approach, can lead to less-biased estimates compared with the case–control approach when the disease is frequent (>1%).

Analyzed parameters

To have the same recall period for cases and controls belonging to the same strata, we excluded from the analysis the thyroid cancers in first-degree relatives when they occurred after the date at which the subject would have the same age as the age at interview of the youngest of the strata. Therefore, in family number 6, both daughters were excluded from the family history of thyroid cancer of their mother, and one of the daughters was excluded from the family history of her sister (see Table 2). A total of 24 cases were therefore considered to have at least one first-degree relative with thyroid cancer for the analysis.

The following variables were constructed: family history of thyroid cancer in a first-degree relative (yes/no), history of thyroid cancer among a female first-degree relative (yes/no), a parent (yes/no), a sibling (yes/no), a mother (yes/no), a sister (yes/no), a daughter (yes/no), a male first-degree relative (yes/no), a father (yes/no), a brother (yes/no), or a son (yes/no).

All ORs were adjusted for ethnicity, the level of education, smoking, interviewer, height, body mass index, head or neck exposure to medical X-ray irradiation before the age of 15, and the number of siblings, and for women, further adjustment for the number of births and menopausal status, which were identified as risk factors in a previous analysis (15,16). Controls were allocated a reference age equal to the case's age at the time of the diagnosis, and only events or exposures that occurred before the reference age were considered for the analysis. Interactions with ethnicity were analyzed. Separated analyses were performed according to the ethnicity and the size of the cancer.

Results

Cases' characteristics by sex are shown in Table 1. Papillary carcinoma was the histological type in 79% of women and 71% of men. The mean difference in the date of birth between each case and his/her control(s) was 46.8 days (standard deviation [SD], 95.5). Among cases, 53% of the women and 50% of the men defined themselves as of Maohi ethnicity. These percentages represented 54% and 64%, respectively, in controls (not shown). Cases had on average of 12.2 first-degree relatives (SD, 5.3), which was significantly higher than controls who had on average of 11 (SD, 4.8) first-degree relatives (p = 0.0042). Median numbers of first-degree relatives were 11 and 10 among cases and controls, respectively.

Table 1.

Characteristics of the Cases of Thyroid Cancer in French Polynesia, 1979–2004

	Women (n = 201)		Men (n = 24)
	n	%	N	%
Histological type
Papillary	158	78.6	17	70.8
Follicular	43	21.4	7	29.2
Size of the tumor (mm)
Papillary
≤4	30	19.0	2	11.8
5–10	51	32.2	5	29.3
11–20	23	14.6	2	11.8
21–40	28	17.7	4	23.5
>40	8	5.1	2	11.8
Microcarcinoma	6	3.8	0	0.0
Unknown	12	7.6	2	11.8
Follicular
≤4	2	4.6	0	0.0
5–10	8	18.6	0	0.0
11–20	3	7.0	0	0.0
21–40	13	30.2	1	14.3
>40	7	16.3	3	42.9
Microcarcinoma	0	0.0	1	14.3
Unknown	10	23.3	2	28.5
Age at diagnosis (years)
10–17	6	3.0	0	0.0
18–29	35	17.4	6	25.0
30–39	69	34.3	7	29.2
40–49	56	27.9	7	29.2
50–55	35	17.4	4	16.6
Year of diagnosis
1979–1985	9	4.5	3	12.5
1986–1990	22	10.9	3	12.5
1991–1995	54	26.9	6	25.0
1996–2000	73	36.3	8	33.3
2001–2004	43	21.4	4	16.7
Multifocal
No	116	57.7	11	45.8
Yes	67	33.3	12	50.0
Missing	18	9.0	1	4.2
Extrathyroid invasion
No	159	79.1	17	70.8
Yes	14	7.0	6	25.0
Missing	28	13.9	1	4.2
Associated thyroid disease or disease extension
Unknown	14	6.9	1	4.1
Goiter	119	59.2	13	54.2
Thyroid nodules	50	24.9	7	29.2
Metastasis	2	1	1	4.2
Lymph node metastasis	2	1	0	0
Thyroid dysfunction	14	7.0	2	8.3

Table 2 describes the 17 families in which at least two individuals with thyroid cancer were identified (cases included in this study and their relatives). In two families (no. 3 and 7), the same histologic diagnosis (papillary) was observed among affected members. In six families, histologic diagnoses among family members were different (one papillary and one follicular in families 2, 4, 5, and 11, and two papillaries and one follicular in family 6). Finally, information on histology was not available for one of the affected family members in the remaining 11 families (1,8 –10,12 –17). The first seven families shown in Table 2 had at least two individuals included as cases in our study. For the analysis, all cases belonging to the same family were attributed a common value for the variable family, whereas all of the other individuals in the study were allocated different values.

Table 2.

Description of the Familial Cases of Thyroid Cancer in French Polynesia, 1979–2004

Family	Sex	Relationship	Age at diagnosis	Size (mm)	Multifocal	Histology
1^a	F	Sister	n.a.	n.a.	n.a.	n.a.
	M	Brother^b	38	15	Yes	Papillary
	F	Sister^b	31	80	No	Papillary
2^a	M	Brother^b	38	30	Yes	Papillary
	F	Sister^b	31	10	Yes	Follicular
3^a	F	Sister^b	26	Micro^c	n.a.	Papillary
	F	Sister^b	46	9	No	Papillary
4^a	M	Father^b	54	Micro^c	No	Follicular
	F	Daughter^b	18	20	No	Papillary
5^a	F	Mother^b	51	50	No	Follicular
	M	Son^b	38	10	Yes	Papillary
6^a	F	Daughter^b	31	10	Yes	Papillary
	F	Daughter^b	24	20	No	Follicular
	F	Mother^b,d	51	32	Yes	Follicular
7^a	M	Father^b	52	5	Yes	Papillary
	F	Daughter^b	44	7	No	Papillary
8	F	Sister	40	n.a.	n.a.	n.a.
	F	Sister^b	51	25	Yes	Papillary
9	F	Mother	71	n.a.	n.a.	n.a.
	F	Daughter^b	53	30	No	Papillary
10	F	Mother	n.a.	n.a.	n.a.	n.a.
	F	Daughter^b	45	n.a.	n.a.	Follicular
11	F	Mother	38	n.a.	Yes	Follicular
	F	Daughter^b	28	n.a.	No	Papillary
12	F	Mother	n.a.	n.a.	n.a.	n.a.
	F	Daughter^b	47	3	No	Papillary
13	F	Sister	n.a.	n.a.	n.a.	n.a.
	F	Sister^b	43	10	No	Papillary
14	F	Sister	71	n.a.	n.a.	n.a.
	F	Sister^b	55	8	Yes	Papillary
15	F	Sister	n.a.	n.a.	n.a.	n.a.
	F	Sister^b	46	n.a.	Yes	Papillary
16	M	Brother	<12	n.a.	n.a.	n.a.
	F	Sister^b	16	30	Yes	Papillary
17	F	Sister	n.a.	n.a.	n.a.	n.a.
	F	Sister^b	71	30	Yes	Follicular

Nonvalidated diagnoses of thyroid cancer are highlighted in gray.

Each of these families from no. 1 to 7 include two cases in the case–control study.

Cases of thyroid cancer included in the study.

Micro indicates a microcarcinoma without precision of its size.

This woman's two daughters were excluded from her family history of thyroid cancer for the analysis and she was not considered as having affected relatives for the analysis.

F, female; M, male; n.a., not available.

Among cases, 21 had one first-degree relative with a thyroid cancer and 3 had two first-degree relatives with thyroid cancer (brother and sister in family no. 1, two sisters in family no. 1, mother and sister in family no. 6). Eleven controls had one first-degree relative with a thyroid cancer. Among first-degree relatives of cases diagnosed with a thyroid cancer, age at diagnosis was less than 30 years for 4 relatives, between 30 and 40 years for 8 relatives, beyond 40 for 11, and unknown for 4. Among first-degree relatives of controls diagnosed with a thyroid cancer, age at diagnosis was less than 30 years for 2, between 30 and 40 years for 3, beyond 40 for 4, and unknown for 2.

ORs of differentiated thyroid cancer associated with a family history of thyroid cancer are presented in Table 3. Individuals who had a first-degree relative with a thyroid cancer had a 4.5-fold increased risk of differentiated thyroid cancer compared with those who did not (95% CI, 1.9–10.6; Table 3). This OR was 4.8 (95% CI, 2.0–11.7) in women and 20.6 (95% CI, 1.2–342) in men. The OR was 10.0 (95% CI, 1.3–74.8) when the relative was a man and 4.0 (95% CI, 1.5–10.3) when the relative was a woman (Table 3). Individuals with a male first-degree relative with thyroid cancer had a nonsignificantly increased risk of differentiated thyroid cancer compared with individuals with a female first-degree relative with thyroid cancer (OR, 3.0; 95% CI, 0.35–26.0). After adjusting on the number of full-term pregnancies and on the menopausal status among women, women with a first-degree relative with thyroid cancer had a 4.8-fold increased risk of differentiated thyroid cancer than those who did not (95% CI, 2.0–11.7; Table 3).

Table 3.

Adjusted Odds Ratios of Thyroid Cancer Associated with a History of Thyroid Cancer Among First-Degree Relatives in French Polynesia, 1979–2004

	Total			Women			Men
Relation to proband	Case/control 225/368	OR ^a	95% CI	Case/control 201/324	OR ^a	95% CI	Case/control 24/44	OR ^a	95% CI
First-degree relative	24/11	4.5	1.9–10.6	19/10	4.8	2.0–11.7	5/1	20.6	1.2–342
Parent	9/4	3.9	1.1–13.8	8/4	4.6	1.3–16.4	1/0	–	–
Sibling	13/5	5.2	1.7–16.0	11/5	5.2	1.6–16.6	2/0	–	–
Children	3/2	2.6	0.2–42.8	1/1	1.3	0.0–157	2/1	15.7	0.7–340
First-degree female relative	20/10	4.0	1.5–10.3	15/9	4.1	1.5–11.4	5/1	20.6	1.2–342
Mother	7/3	3.9	0.8–18.2	6/3	4.4	0.9–21.5	1/0	–	–
Sister	11/5	4.2	1.2–14.0	9/5	4.2	1.2–14.9	2/0	–	–
Daughter	3/2	2.6	0.2–42.8	1/1	1.3	0.0–157	2/1	15.7	0.7–340
First-degree male relative	5/1	10.0	1.3–74.8	5/1	9.9	1.6–62.8	0/0	–	–
Father	2/1	3.5	0.4–29.1	2/1	4.5	0.6–35.4	0/0	–	–
Brother	3/0	–	–	3/0	–	–	0/0	–	–

Reference group in each analysis consists of subjects with no history of thyroid cancer in that relative.

Adjusted for ethnicity, level of education, smoking, interviewer, height, body mass index, head or neck exposure to medical X-ray irradiation before the age of 15, and number of siblings, and for women, further adjustment for the number of births and menopausal status.

OR, odds ratio; CI, confidence interval.

The separate analyses according to ethnicity showed that among Maohis, individuals who had a first-degree relative with thyroid cancer had an 11-fold increased risk of differentiated thyroid cancer (95% CI, 2.4–53.6) compared with those who did not, whereas this risk was 2.1 (95% CI, 0.7–5.8) among individuals of mixed origin (p for interaction between ethnicity and familial history of thyroid cancer: 0.07) (results not shown).

When restricting the analysis to the strata of cases with microcarcinomas (carcinomas measuring less than 10 mm, excluding those with associated metastasis, lymph node or extravascular invasion, and multifocal cancers), five cases and five controls had a first-degree relative with a thyroid cancer. In this subgroup, a family history of thyroid cancer was associated with a nonsignificant threefold increased risk of differentiated thyroid cancer (95% CI, 0.6–13.8; p = 0.2). Among cancers larger than 10 mm (carcinomas measuring more than 10 mm and carcinomas measuring less than 10 mm but associated with metastasis, lymph nodes or extravascular invasion, or multifocal carcinomas), 12 cases and 4 controls had a first-degree relative with a thyroid cancer. In this subgroup, a family history of thyroid cancer was associated with a sixfold increased risk of differentiated thyroid cancer (95% CI, 1.8–20.5; p = 0.004) (results not shown).

Discussion

In our case–control study of differentiated thyroid cancer in French Polynesia, individuals who had a family history of thyroid cancer in a first-degree relative had a 4.5-fold increased risk of differentiated thyroid cancer compared with those who did not. The risk increased 10-fold when the concerned relative was a man and increased 4-fold when the relative was a woman. This association was significant only for cancers measuring 10 mm or more, but not for cancers measuring less than 10 mm. Moreover, the analysis according to ethnicity showed that the association was borderline significantly (p = 0.07) higher among individuals of Maohi origin than among individuals of mixed origin.

We were able to validate most of the thyroid cancers that occurred among first-degree relatives in our study. This was important because any nonmalignant disease, erroneously declared as thyroid cancer, such as a multinodular goiter in one individual (a mother), could be excluded. Only five thyroid cancers among first-degree relatives of cases and two thyroid cancers in first-degree relatives of controls were not validated.

As cases had to be less than 56, we had less information on their family members than if we also had older cases, for which there would be a longer lifespan to study the outcomes of their first-degree relatives. The power of our study of the association between family history of thyroid cancer and risk of differentiated thyroid cancer may have been slightly reduced by this cutoff choice. The eligible cases excluded from our analysis because they were dead (n = 14) or too ill (n = 1) to be interviewed could correspond to particular clinical forms (more severe thyroid cancer or diagnosis at a later stage of the disease). Their exclusion may have induced a survival bias in the following manner: either these forms were related to familial forms, leading to an underestimation of the association that we observed between a family history of thyroid cancer and the risk of differentiated thyroid cancer, or they could also correspond to sporadic forms, and excluding them may then have overestimated the association. ORs of differentiated thyroid cancers associated with a family history of thyroid cancer among first-degree relatives could have been overestimated because of a memory bias: cases were able to remember their family history of thyroid cancer better than controls. In addition, in our study, the diagnosis of thyroid cancer among family members sometimes occurred after the case's diagnosis. However, the resulting screening bias is minimal because the families of the cases did not systematically receive a more thorough medical examination in search of thyroid cancers than families of controls.

The risk of thyroid cancer associated with a family history of thyroid cancer was studied in four case–control studies (10 –13,21). Although each of the studies included from 159 (13) to 608 cases (10), only the largest one observed a significantly increased risk of thyroid cancer when there was a positive family history of thyroid cancer, considering all degrees of relatives (OR, 6.1; 95% CI, 1.8–21.3) (10). Our finding of an OR of 4.5 for a first-degree relative is consistent with the results from other studies. Although those results were not statistically significant, ORs attained 7.8 (95% CI, 0.5–112.8) for a first-degree relative (11), 5.2 (95% CI, 0.2–554) for a first-degree relative or a grandparent (13,21), and 3.0 (95% CI, 0.8–11.1) for a first, second, or third-degree relative (12). Our finding was significant, but the result of the other case–control studies of similar size was not. This could be attributed to a greater power due to a higher incidence of thyroid cancer and/or larger families.

The results obtained from studies based on cancer registries in Sweden (8,9,22), Norway (23), and Utah (24) corroborate these findings. In Sweden, the study was based on the nationwide Family-Cancer Database of 9.6 million individuals including children born after 1934 and their parents (9). Among all the malignancies studied, the FHR was the highest for thyroid cancer (standardized incidence ratio [SIR] of thyroid cancer among children of parents with thyroid cancer over SIR among children of parents without thyroid cancer, FHR, 10.7; 95% CI, 6.9–16.6) (8).

Studies with cohort approaches also corroborate these results. They evaluated the risk of developing a thyroid cancer among relatives of individuals with a thyroid cancer. Two studies estimated the risk of thyroid cancer among the parents or first-degree relatives of individuals with a thyroid cancer (25,26). In the first study, parents of individuals with thyroid cancer exhibited a fourfold increased risk of thyroid cancer compared with the general population of the same age, sex, and birth cohort distribution (25). In another study, the incidence of thyroid cancer was 10-fold higher among first-degree relatives of cases (incidence rate ratio, 10.3; 95% CI, 2.2–47.6) than among first-degree relatives of controls (26). Much larger studies based on data from cancer registries in Utah (24) and Norway (23) reported an elevated incidence of thyroid cancer among first-degree relatives of individuals with thyroid cancer.

Finally, in the Swedish nationwide Family-Cancer Database, with a follow-up period spanning 16 years (1986–2002), which focused on nonmedullary thyroid cancers, the SIR of thyroid adenocarcinoma when a sibling had any thyroid cancer was 5.15 (95% CI, 2.80–8.66). There was only one brother–sister pair, and all others were sister–sister pairs. The risks were 1.48 (95% CI, 0.00–8.47) when a brother was the proband and 6.36 (95% CI, 3.37–10.90) when a sister was the proband (27).

Differentiated thyroid carcinoma heritability seems to be higher among men than among women. In the Swedish registry-based study, the sons of parents with a papillary or a follicular thyroid cancer had a 7.8-fold higher incidence than men of the same age in the general population (95% CI, 3.9–13.2). This ratio was similar when either the mother (SIR, 7.7; 95% CI, 3.3–14.0) or the father (SIR, 8.2; 95% CI, 1.5–20.0) was affected. This ratio was lower in daughters of affected parents (SIR, 2.8; 95% CI, 1.5–4.5), but it was not significantly higher for mothers (SIR, 3.5; 95% CI, 1.8–5.7), than for fathers (SIR, 0.8; 95% CI, 0.0–3.2) (9). Because of the small number of male cases of differentiated thyroid cancer in our study, our results are exclusively informative for women. In our study, individuals who had an affected parent had a 3.9-fold increased risk of differentiated thyroid cancer compared with individuals without an affected parent (95% CI, 1.1–13.8), and the OR was 4.6 (95% CI, 1.3–16.4) for women.

In the study based on the Swedish nationwide Family-Cancer Database, offspring (either male or female) whose mother or father was diagnosed with nonmedullary thyroid cancer had a 4.50 (95% CI, 2.74–6.96) and a 2.21 (95% CI, 0.57–5.70) increased risk of adenocarcinoma of the thyroid gland (26). Risks were equally high for sons and daughters when the mother was the proband. These results are also consistent with our study, where these risks were 3.9 (95% CI, 0.8–18.2) when the mother was diagnosed and 3.5 (95% CI, 0.4–29.1) when the father was diagnosed with thyroid cancer.

Overall, in our study, individuals with a male first-degree relative diagnosed with thyroid cancer had a nonsignificantly increased risk of differentiated thyroid cancer compared with individuals whose affected relative was a female first-degree relative. The four other case–control studies that analyzed an association with a family history of thyroid cancer did not report results according to the sex of the affected relative, mostly because the number of relatives was smaller (10 –13,21).

In our study, familial inheritance was nonsignificantly higher in Maohis than in the other ethnic group. Although this result was not significant, it could be related to the fact that the incidence of thyroid cancer is higher among Maohis than among immigrants of other origins living in French Polynesia (2).

Some studies suggest that nonmedullary thyroid cancer is inherited in an autosomal dominant fashion with incomplete penetrance (28,29). However, other models of hereditary involving an elevated number of different genes could also be compatible (28,29). For the moment, the genetic abnormalities linked to familial predisposition are unknown (29).

Family adenomatous polyposis (FAP) and Cowden syndrome, two highly penetrant hereditary cancer syndromes, but with a low incidence (1/5000–10,000 for FAP and 1/200,000 for Cowden syndrome), are autosomal dominant conditions (mutations in the APC gene for FAP and in the PTEN gene for Cowden syndrome) that give rise to an increased incidence of thyroid cancer (30). About 2% of individuals with FAP will develop a thyroid cancer during their lifetime and about 3–10% affected by Cowden syndrome (30). However, because of their rarity, these syndromes are unlikely to explain the aggregation of thyroid cancers in our study. Moreover, the identification and the estimation of the prevalence of those syndromes in thyroid cancer cohorts is impossible because of the lack of information on associated pathologies (such as hamartomas in Cowden syndrome or colic polyps in the FAP) among subjects in these large cohort studies. A small case–control study like ours is unable to investigate such syndromes.

Until now, only candidate gene approaches have been used to investigate the association between DNA polymorphism and differentiated thyroid cancer risk. Around 30 different genes were investigated in a similar number of studies (31). In a recent large study, the GC heterozygous state of pre-miRNA-146a was found to be strongly associated with an increased risk of differentiated thyroid carcinoma (OR, 1.6; 95% CI, 1.3–2.0) (32). Among the other studies that were smaller, the majority of the results were found in only one study and need to be confirmed (31). Only two findings were confirmed in more than one study: a fairly high risk was evidenced for GSTM1- and GSTT1-null genotypes in three studies, without heterogeneity between the studies (33 –35). In addition, an increased risk for homozygous carriers of the P53 Arg72/pro allele was found in the two studies that investigated this gene. The pooled OR was 2.0 (95% CI, 1.2–3.6), but with heterogeneity due to the high proportion of homozygous carriers among controls in one study (36,37).

In conclusion, our study confirms that the familial inheritance of differentiated thyroid cancer is high and shows that it is particularly high in Maohi populations.

Author Contribution Statement

Pauline Brindel participated in the management of the study data, conducted the analysis, and participated in the interpretation of results and the writing of the article. Florent de Vathaire and Françoise Doyon, as principal investigators, elaborated the study design, obtained an ethical approval, and were involved in the analysis and interpretation of results and the writing of the article. Catherine Bourgain participated in the analysis and interpretation of results. Frédérique Rachedi, Jean-Louis Boissin, Joseph Sebbag, and Larrys Shan participated in the organization of the study and the recruitment of the cases. Frédérique Bost-Bezeaud and Patrick Petitdidier confirmed the pathological diagnoses. John Paoaafaite and Joseph Teuri contacted and interviewed cases and controls.

Footnotes

Acknowledgments

This work was carried out at Unit 1018, INSERM, 39 rue Camille Desmoulins, 94805 Villejuif Cedex, France. The authors are grateful to Ms. Lorna Saint-Ange for editing. The authors thank C. Bonaïti for methodological help, and A. Klouman, C. Rougeolles, P. Morales, J. Iltis, P. Giraud, P. Didiergeorge, M. Brisard, G. Soubiran, B. Caillou, J.M. Bidard, A. Merceron, M.L. Vanizette, P. Dupire, M. Berges, J. Ienfa, G. de Clermont, N. Cerf, B. Oddo, M. Bambridge, C. Baron, A. Mouchard-Rachet, O. Simonet, D. Lamarque, J. Vabret, J. Delacre, M.P. Darquier, and J. Leninger for their help in the collection of the cases or in the organization of the case-control study. This study was supported by the Ligue Nationale Contre le Cancer (LNCC), the Direction Générale de la Santé (DGS), the Comité de Radioprotection de l'EDF, the Agence Française de Sécurité Sanitaire et Environnementale (AFSSE), and the CHILD-THYR EEC program. Pauline Brindel was supported by grants from the ARC and the Institut National du Cancer.

Disclosure Statement

There is no conflict of interest that could be perceived as prejudicing the impartiality of the research reported.

References

de Vathaire

, Le Vu

, Vathaire

. 2000. Thyroid cancer in French Polynesia between 1985 and 1995: influence of atmospheric nuclear bomb tests performed at Mururoa and Fangataufa between 1966 and 1974. Cancer Causes Control, 11:59–63.

Gleize

, Laudon

, Sun

, Challeton-de Vathaire

, Le Vu

, de Vathaire

. 2000. Cancer registry of French Polynesia: results for the 1990–1995 period among native and immigrant population. Eur J Epidemiol, 16:661–667.

Cardis

, Kesminiene

, Ivanov

, Malakhova

, Shibata

, Khrouch

, Drozdovitch

, Maceika

, Zvonova

, Vlassov

, Bouville

, Goulko

, Hoshi

, Abrosimov

, Anoshko

, Astakhova

, Chekin

, Demidchik

, Galanti

, Ito

, Korobova

, Lushnikov

, Maksioutov

, Masyakin

, Nerovnia

, Parshin

, Parshkov

, Piliptsevich

, Pinchera

, Polyakov

, Shabeka

, Suonio

, Tenet

, Tsyb

, Yamashita

, Williams

. 2009. Risk of thyroid cancer after exposure to 131I in childhood. J Natl Cancer Inst, 97:724–732.

De Vathaire

, Hardiman

, Shamsaldin

, Campbell

, Grimaud

, Hawkins

, Raquin

, Oberlin

, Diallo

, Zucker

, Panis

, Lagrange

, Daly-Schveitzer

, Lemerle

, Chavaudra

, Schlumberger

, Bonaïti

. 1999. Thyroid carcinomas after irradiation for a first cancer during childhood. Arch Intern Med, 159:2713–2719.

Lindberg

, Karlsson

, Arvidsson

, Holmberg

, Lunberg

, Wallgren

. 1995. Cancer incidence after radiotherapy for skin haemangioma during infancy. Acta Oncol, 34:735–740.

Lundell

, Hakulinen

, Holm

. 1994. Thyroid cancer after radiotherapy for skin hemangioma in infancy. Radiat Res, 140:334–339.

Ron

, Lubin

, Shore

, Mabuchi

, Modan

, Pottern

, Schneider

, Tucker

, Boice

. 1995. Thyroid cancer after exposure to external radiation: a pooled analysis of seven studies. Radiat Res, 141:259–277.

Hemminki

, Vaittinen

, Kyyronen

. 1998. Age-specific familial risks in common cancers of the offspring. Int J Cancer, 78:172–175.

Hemminki

, Dong

. 2000. Familial relationships in thyroid cancer by histo-pathological type. Int J Cancer, 85:201–205.

10.

Horn-Ross

, Morris

, Lee

, West

, Whittemore

, McDougall

, Nowels

, Stewart

, Spate

, Shiau

, Krone

. 2001. Iodine and thyroid cancer risk among women in a multiethnic population: the Bay Area Thyroid Cancer Study. Cancer Epidemiol Biomark Prev, 10:979–985.

11.

Mack

, Preston-Martin

, Bernstein

, Qian

. 2002. Lifestyle and other risk factors for thyroid cancer in Los Angeles County females. Ann Epidemiol, 12:395–401.

12.

Memon

, Berrington

, Luqmani

, Suresh

. 2004. Family history of benign thyroid disease and cancer and risk of thyroid cancer. Eur J Cancer, 40:754–760.

13.

Ron

, Kleinerman

, Boice

Jr. , LiVolsi

, Flannery

, Fraumeni

. 1987. A population-based case-control study of thyroid cancer. J Natl Cancer Inst, 79:1–12.

14.

United Nations Scientific Committee on the Effects of Atomic R. 2000. Sources and Effects of Ionizing Radiation. Report to the General Assembly with Annexes: New York: United Nations.

15.

Brindel

, Doyon

, Rachedi

, Boissin

, Sebbag

, Shan

, Chungue

, Bost-Bezeaud

, Petitdidier

, Paoaafaite

, Teuri

, de Vathaire

. 2008. Anthropometric factors in differentiated thyroid cancer in French Polynesia: a case-control study. Cancer Causes Control, 20:581–590.

16.

Brindel

, Doyon

, Rachedi

, Boissin

, Sebbag

, Shan

, Chung

, Kai Sun

, Bost-Bezeaud

, Petitdidier

, Morales

, Paoaafaite

, Teuri

, de Vathaire

. 2008. Menstrual and reproductive factors in the risk of differentiated thyroid carcinoma in native women in French Polynesia: a population-based case-control study. Am J Epidemiol, 167:219–229.

17.

Perdu

, Boutouyrie

, Bourgain

, Stern

, Laloux

, Bozec

, Azizi

, Bonaiti-Pellié

, Plouin

, Laurent

, Gimenez-Roqueplo

, Jeunemaitre

. 2007. Inheritance of arterial lesions in renal fibromuscular dysplasia. J Hum Hypertens, 21:393–400.

18.

Kerber

. 1995. Method for calculating risk associated with family history of a disease. Genet Epidemiol, 12:291–301.

19.

Malecot

. 1948. Les Mathématiques de l'Hérédité Masson . Paris.

20.

Khoury

, Flanders

. 1995. Bias in using family history as a risk factor in case-control studies of disease. Epidemiology, 6:511–519.

21.

Ron

, Kleinerman

, LiVolsi

, Fraumeni

. 1991. Familial nonmedullary thyroid cancer. Oncology, 48:309–311.

22.

Hemminki

, Vaittinen

. 1997. Effect of paternal and maternal cancer on cancer in the offspring: a population-based study. Cancer Epidemiol Biomark Prev, 6:993–997.

23.

Frich

, Glattre

, Akslen

. 2001. Familial occurrence of nonmedullary thyroid cancer: a population-based study of 5673 first-degree relatives of thyroid cancer patients from Norway. Cancer Epidemiol Biomark Prev, 10:113–117.

24.

Goldgar

, Easton

, Cannon-Albright

, Skolnick

. 1994. Systematic population-based assessment of cancer risk in first-degree relatives of cancer probands. J Natl Cancer Inst, 86:1600–1608.

25.

Galanti

, Ekbom

, Grimelius

, Yuen

. 1997. Parental cancer and risk of papillary and follicular thyroid carcinoma. Br J Cancer, 75:451–456.

26.

Pal

, Vogl

, Chappuis

, Tsang

, Brierley

, Renard

, Sanders

, Kantemiroff

, Bagha

, Goldgar

, Narod

, Foulkes

. 2001. Increased risk for nonmedullary thyroid cancer in the first degree relatives of prevalent cases of nonmedullary thyroid cancer: a hospital-based study. J Clin Endocrinol Metab, 86:5307–5312.

27.

Hemminki

, Eng

, Chen

. 2005. Familial risks for nonmedullary thyroid cancer. J Clin Endocrinol Metab, 90:5747–5753.

28.

Lesueur

, Stark

, Tocco

, Ayadi

, Delisle

, Goldgar

, Schlumberger

, Romeo

, Canzian

. 1999. Genetic heterogeneity in familial nonmedullary thyroid carcinoma: exclusion of linkage to RET, MNG1, and TCO in 56 families. NMTC Consortium. J Clin Endocrinol Metab, 84:2157–2162.

29.

Sturgeon

, Clark

. 2005. Familial nonmedullary thyroid cancer. Thyroid, 15:588–593.

30.

Nagy

, Sweet

, Eng

. 2004. Highly penetrant hereditary cancer syndromes. Oncogene, 23:6445–6670.

31.

Adjadj

, Schlumberger

, de Vathaire

. 2009. Germ-line DNA polymorphisms and susceptibility to differentiated thyroid cancer. Lancet Oncol, 10:181–190.

32.

Jazdzewski

, Murray

, Franssila

, Jarzab

, Schoenberg

, de la Chapelle

. 2008. Common SNP in pre-miR-146a decreases mature miR expression and predisposes to papillary thyroid carcinoma. Proc Natl Acad Sci USA, 105:7269–7274.

33.

Hernández

, Céspedes

, Xamena

, Surrallés

, Creus

, Galofré

, Marcos

. 2003. Glutathione S-transferase polymorphisms in thyroid cancer patients. Cancer Lett, 190:37–44.

34.

, Zhao

, Zheng

, Liu

, Wei

, Sturgis

. 2008. Glutathione S-transferase polymorphisms and risk of differentiated thyroid carcinomas: a case-control analysis. Arch Otolaryngol Head Neck Surg, 132:756–761.

35.

Morari

, Leite

, Granja

, da Assumpcao

, Ward

. 2002. The null genotype of glutathione S-transferase M1 and T1 locus increases the risk for thyroid cancer. Cancer Epidemiol Biomark Prev, 11:1485–1488.

36.

Granja

, Morari

, Correa

, Assumpcao

, Ward

. 2004. Proline homozygosity in codon 72 of p53 is a factor of susceptibility for thyroid cancer. Cancer Lett, 210:151–157.

37.

Rogounovitch

, Saenko

, Ashizawa

, Sedliarou

, Namba

, Abrosimov

, Lushnikov

, Roumiantsev

, Konova

, Petoukhova

, Tchebotareva

, Ivanov

, Chekin

, Bogdanova

, Tronko

, Tsyb

, Thomas

, Yamashita

. 2006. TP53 codon 72 polymorphism in radiation-associated human papillary thyroid cancer. Oncol Rep, 15:949–956.