Progression of BRAF-induced thyroid cancer is associated with epithelial-mesenchymal transition requiring concomitant MAP kinase and TGFß signaling.

Mice with thyroid-specific expression of oncogenic BRAF (Tg-Braf) develop papillary thyroid cancers (PTCs) that are locally invasive and have well-defined foci of poorly differentiated thyroid carcinoma (PDTC). To investigate the PTC-PDTC progression, we performed a microarray analysis using RNA from paired samples of PDTC and PTC collected from the same animals by laser capture microdissection. Analysis of eight paired samples revealed a profound deregulation of genes involved in cell adhesion and intracellular junctions, with changes consistent with an epithelial-mesenchymal transition (EMT). This was confirmed by immunohistochemistry, as vimentin expression was increased and E-cadherin lost in PDTC compared with adjacent PTC. Moreover, PDTC stained positively for phospho-Smad2, suggesting a role for transforming growth factor (TGF)ß in mediating this process. Accordingly, TGFß-induced EMT in primary cultures of thyroid cells from Tg-Braf mice, whereas wild-type thyroid cells retained their epithelial features. TGFß-induced Smad2 phosphorylation, transcriptional activity and induction of EMT required mitogen-activated protein kinase (MAPK) pathway activation in Tg-Braf thyrocytes. Hence, tumor initiation by oncogenic BRAF renders thyroid cells susceptible to TGFß-induced EMT, through a MAPK-dependent process.

DNA breaks at fragile sites generate oncogenic RET/PTC rearrangements in human thyroid cells.

Human chromosomal fragile sites are regions of the genome that are prone to DNA breakage, and are classified as common or rare, depending on their frequency in the population. Common fragile sites frequently coincide with the location of genes involved in carcinogenic chromosomal translocations, suggesting their role in cancer formation. However, there has been no direct evidence linking breakage at fragile sites to the formation of a cancer-specific translocation. Here, we studied the involvement of fragile sites in the formation of RET/PTC rearrangements, which are frequently found in papillary thyroid carcinoma (PTC). These rearrangements are commonly associated with radiation exposure; however, most of the tumors found in adults are not linked to radiation. In this study, we provide structural and biochemical evidence that the RET, CCDC6 and NCOA4 genes participating in two major types of RET/PTC rearrangements, are located in common fragile sites FRA10C and FRA10G, and undergo DNA breakage after exposure to fragile site-inducing chemicals. Moreover, exposure of human thyroid cells to these chemicals results in the formation of cancer-specific RET/PTC rearrangements. These results provide the direct evidence for the involvement of chromosomal fragile sites in the generation of cancer-specific rearrangements in human cells.

Interphase chromosome folding determines spatial proximity of genes participating in carcinogenic RET/PTC rearrangements.

Recurrent chromosomal rearrangements are common in cancer cells and may be influenced by nonrandom positioning of recombination-prone genetic loci in the nucleus. However, the mechanism responsible for spatial proximity of specific loci is unknown. In this study, we use an 18 Mb region on 10q11.2-21 containing the RET gene and its recombination partners, the H4 and NCOA4 (ELE1) genes, as a model chromosomal region frequently involved in RET/PTC rearrangements in thyroid cancer. RET/PTC is particularly common in tumors from children exposed to ionizing radiation. Using fluorescence in situ hybridization and three-dimensional microscopy, the locations of five different loci in this region were mapped in interphase nuclei of normal human thyroid cells. We show that RET and NCOA4 are much closer to each other than expected based on their genomic separation. Modeling of chromosome folding in this region suggests the presence of chromosome coiling with coils of approximately 8 Mb in length, which positions the RET gene close to both, the NCOA4 and H4, loci. There was no significant variation in gene proximity between adult and pediatric thyroid cells. This study provides evidence for large-scale chromosome folding of the 10q11.2-21 region that offers a structural basis for nonrandom positioning and spatial proximity of potentially recombinogenic intrachromosomal loci.

Solid variant of papillary thyroid carcinoma: incidence, clinical-pathologic characteristics, molecular analysis, and biologic behavior.

Solid variant is a rare and poorly characterized variant of papillary thyroid carcinoma. In this study we analyzed 20 primary cases of the solid variant of papillary carcinoma found in a series of 756 papillary carcinomas operated at the Mayo Clinic between 1962 and 1989. The criteria for classification included predominantly (>70%) solid growth pattern of primary tumor, retention of cytologic features typical of papillary carcinoma, and absence of tumor necrosis. For each case of the solid variant, a control case of classical papillary carcinoma matched by age, sex, tumor size, and length of follow-up was selected. The follow-up ranged from 6 to 32 years. Two patients with the solid variant of papillary carcinoma (10%) died from disease 7 and 10 years after initial surgery, while another two patients (10%) are alive with lung metastases. In contrast, the control group had no cases with distant metastases or death from disease. Molecular analyses showed a similar prevalence of RET /PTC rearrangements in both groups. In conclusion, the solid variant of papillary carcinoma is associated with a slightly higher frequency of distant metastases and less favorable prognosis than classical papillary carcinoma. However, it should be distinguished from poorly differentiated thyroid carcinoma, which has a reported lower survival rate compared with the solid variant of papillary carcinoma.

Spatial positioning of RET and H4 following radiation exposure leads to tumor development.

Exposure to ionizing radiation is a well-known risk factor for a number of human cancers, including leukemia, thyroid cancer, soft tissue sarcomas, and many others. Although it has been known for a long time that radiation exposure to the cell results in extensive DNA damage, including double strand DNA breaks, the exact mechanisms of radiation-induced carcinogenesis remain unknown. Recently, a large increase in incidence of thyroid cancer was observed in children exposed to radiation after the Chernobyl nuclear accident. A high prevalence of chromosomal rearrangements involving the RET gene was found among these radiation-induced thyroid tumors. As a result of such rearrangement, a portion of the RET gene is fused with another gene, typically with the H4 or ELE1 . However, since the DNA targets of ionizing radiation are randomly distributed throughout the cell nucleus, the reason for predilection for the RET rearrangements in thyroid cells was unclear.

On target cell numbers in radiation-induced H4-RET mediated papillary thyroid cancer.

Radiation-induced human papillary thyroid cancer (PTC) is associated with chromosomal inversions that involve the genetic loci H4 and RET on chromosome 10. Recently, experimental data has shown that these loci lie in very close spatial proximity in a high proportion of adult human thyroid cells. Applying the generalized formulation of dual radiation action to this H4-to-RET geometric distance data, we predict here the radiation dose-response of H4-RET induction. The predicted H4-RET dose-response has a linear-to-quadratic transition dose of approximately 7 Gy, suggesting the validity of linear risk extrapolations to very low doses for H4-RET mediated radiation-induced PTC. In conjunction with A-bomb survivor data, the predicted H4-RET dose-response yields estimates of the number of PTC target cells that are of the order of approximately 10(6) to approximately 10(7) cells, i.e. considerably less than the total number of follicular cells in the thyroid gland.

Proximity of chromosomal loci that participate in radiation-induced rearrangements in human cells.

Rearrangements involving the RET gene are common in radiation-associated papillary thyroid cancer (PTC). The RET/PTC1 type of rearrangement is an inversion of chromosome 10 mediated by illegitimate recombination between the RET and the H4 genes, which are 30 megabases apart. Here we ask whether despite the great linear distance between them, RET and H4 recombination might be promoted by their proximity in the nucleus. We used two-color fluorescence in situ hybridization and three-dimensional microscopy to map the positions of the RET and H4 loci within interphase nuclei. At least one pair of RET and H4 was juxtaposed in 35% of normal human thyroid cells and in 21% of peripheral blood lymphocytes, but only in 6% of normal mammary epithelial cells. Spatial contiguity of RET and H4 may provide a structural basis for generation of RET/PTC1 rearrangement by allowing a single radiation track to produce a double-strand break in each gene at the same site in the nucleus.

Thyroid transcription factor-1, thyroglobulin, cytokeratin 7, and cytokeratin 20 in thyroid neoplasms.

Thyroid transcription factor-1 (TTF-1), a member of the NKx2 family of homeodomain transcription factors, is a mediator of thyroid-specific transcription of the thyroglobulin (TG) gene. The combined immunohistochemical profile of TTF-1, TG, cytokeratin 7 (CK7), and cytokeratin 20 (CK20) in neoplasms of the thyroid gland and their metastases to other sites has not been defined previously. Formalin-fixed tissue of 43 thyroid tumors, including 31 carcinomas and 12 adenomas, and 16 metastasic lesions were immunostained using monoclonal antibodies to TTF-1, TG, CK7, and CK20. Immunoreactivity of the primary tumors (adenomas and carcinomas) for TTF-1 was seen in 32 cases (74%), TG 32 (74%), and CK7 34 (79%), whereas none (0%) showed positivity for CK20. The distribution of reactivity in the 31 carcinomas for TTF-1, TG, and CK7, respectively was papillary (8/8), (8/8), and (8/8); poorly differentiated (6/7), (4/7), and (6/7); oncocytic (Hürthle) cell (2/6), (6/6), and (4/6); follicular (4/4), (3/4), and (3/4); medullary (1/2), (0/2), and (1/2). One of four anaplastic carcinomas was focally immunoreactive showing positivity for TTF-1 only. Of the six follicular adenomas, five were positive for TTF-1, six for TG, and six for CK7. Among the six oncocytic cell adenomas, five were reactive for TTF-1, five for TG, and all six for CK7. Twelve (75%) of the 16 metastatic tumors were positive for TTF-1, 10 (63%) for TG, 15 (94%) for CK7, and none (0%) for CK20. In summary, TTF-1 and TG are demonstrable by immunohistochemistry in the majority of thyroid neoplasms. Compared with TG, an antibody to TTF-I is a similarly sensitive marker for thyroid tumors. Moreover, TTF-1 is a more sensitive marker for poorly differentiated carcinomas and metastasis. In most cases, its nuclear pattern of immunoreactivity facilitates interpretation. Thyroid tumors are CK7+/CK20-. The panel of antibodies for TG, TTF-1, CK7, and CK20 is useful when the thyroid origin of a metastatic tumor is a consideration.

Tyrosine kinase expression is increased in papillary thyroid carcinoma of children and young adults.

Tyrosine kinases (TKs) are important candidate genes for malignant transformation and at least 21 different TKs have been identified in the thyroid gland. We hypothesized that the collective activity of these TKs might be increased in thyroid carcinoma and have association with the clinical behavior of individual tumors. To test this, we determined TK expression by immunohistochemistry in 74 archival thyroid tissue blocks (48 papillary thyroid carcinoma, PTC; 9 follicular thyroid carcinoma, FTC; 17 benign thyroid diseases) from children and young adults. Mean TK expression was greater for PTC (2.1 +/- 0.11) than benign lesions (1.6 +/- 0.2, p = 0.027), and also tended to be greater in FTC (2.1 +/- 0.25, p = 0.12). Recurrence risk was three-fold greater for PTC with intense TK expression (4/15, 27%) than for PTC with minimal - moderate TK expression (3/33, 9.0%). However, this was not statistically significant (p = 0.10). In PTC, TK expression correlated with expression of the receptor for hepatocyte growth factor / scatter factor (cMET, r = 0.31, p = 0.044). In FTC, TK expression did not correlate with cMET, but tended to be greater in young patients (r = -0.59, p = 0.09). We conclude that TK expression is increased in PTC and possibly associated with an increased recurrence risk.

Chromosomal breakpoint positions suggest a direct role for radiation in inducing illegitimate recombination between the ELE1 and RET genes in radiation-induced thyroid carcinomas.

The RET/PTC3 rearrangement is formed by fusion of the ELE1 and RET genes, and is highly prevalent in radiation-induced post-Chernobyl papillary thyroid carcinomas. We characterized the breakpoints in the ELE1 and RET genes in 12 post-Chernobyl pediatric papillary carcinomas with known RET/PTC3 rearrangement. We found that the breakpoints within each intron were distributed in a relatively random fashion, except for clustering in the Alu regions of ELE1. None of the breakpoints occurred at the same base or within a similar sequence. There was also no evidence of preferential cleavage in AT-rich regions or other target DNA sites implicated in illegitimate recombination in mammalian cells. Modification of sequences at the cleavage sites was minimal, typically involving a 1-3 nucleotide deletion and/or duplication. Surprisingly, the alignment of ELE1 and RET introns in opposite orientation revealed that in each tumor the position of the break in one gene corresponded to the position of the break in the other gene. This tendency suggests that the two genes may lie next to each other but point in opposite directions in the nucleus. Such a structure would facilitate formation of RET/PTC3 rearrangements because a single radiation track could produce concerted breaks in both genes, leading to inversion due to reciprocal exchange via end-joining.

Frequent loss of heterozygosity at chromosome 3p14.2-3p21 in human pancreatic islet cell tumours.

OBJECTIVE: Pancreatic islet betacell tumours occur either sporadically or as part of inherited neoplastic syndromes, most commonly multiple endocrine neoplasia (MEN) type 1. Recently, a transgenic mouse model has been established in which the expression of the SV40 large T antigen was targeted to betacells by the rat insulin promoter, leading to the development of multiple pancreatic betacell tumours. In the advanced stages of tumour evolution, these tumours exhibited a high prevalence of loss of heterozygosity (LOH) on mouse chromosomes 9 and 16, at regions syntenic with regions 3q, 3p21, 6q12, 15q24 and 22q of the human genome. DESIGN: Loss of heterozygosity in human islet cell tumours was analysed in a PCR based approach at regions of the human genome syntenic with the mouse loci linked to pancreatic betacell tumours as well as the MEN1 gene on chromosome 11q13. These included 35 microsatellite markers in the human chromosomal regions 3q, 3p21, 6q12, 11q13, 15q24 and 22q. PATIENTS: 21 patients diagnosed with insulinoma were analysed. Histologically, 16 tumours were benign, while 5 were malignant insulinomas. RESULTS: Thirteen of 21 (62%) tumours were found to have loss of genetic material on chromosome 3. The shortest region of overlap implicated a deletion at 3p14.2-3p21 region, corresponding to the marker D3S1295. We did not detect a substantial frequency of LOH in the other syntenic regions, except for the region of MEN 1 gene on 11q13 found to be deleted in 6 (29%) cases, including 3 of 4 tumours from MEN 1 families. Deletions of 3p14. 2-3p21 were observed in 8 of 15 (53%) benign tumours, and in 5 of 6 (83%) malignant neoplasms. CONCLUSIONS: These results indicate the high frequency of 3p14.2-3p21 deletions in human pancreatic betacell neoplasms. These finding suggest the presence of a tumour suppressor gene in this region, that may be important in the microevolution of these tumours towards malignancy.

Pathomorphology of thyroid gland lesions associated with radiation exposure: the Chernobyl experience and review of the literature.

The common sources of ionizing radiation exposure to the thyroid gland in humans are accidental environmental exposure and medical, therapeutic, or diagnostic irradiation. Radiation often induces notable histologic changes in the thyroid tissue and is a well-established risk factor for benign and malignant thyroid tumors. In this paper, we review the acute and chronic histologic changes in the thyroid gland subjected to irradiation, and characterize benign thyroid nodules and malignant tumors arising after exposure, with particular emphasis on thyroid lesions in the population exposed to ionizing radiation as a result of the Chernobyl nuclear accident.

Prevalence of minisatellite and microsatellite instability in radiation-induced post-Chernobyl pediatric thyroid carcinomas.

Exposure to ionizing radiation induces different forms of genomic instability in cultured cells and experimental animals. A higher rate of germline mutations at human hypervariable minisatellite loci was reported in children born from parents exposed to radiation after Chernobyl, implicating genome destabilization as a possible mechanism responsible for late radiation effects in humans. To test if radiation-induced carcinogenesis in the thyroid gland may be associated with somatic minisatellite instability or microsatellite instability, we utilized a PCR-based approach to study normal and tumor DNA from 17 pediatric post-Chernobyl papillary thyroid carcinomas for mutations at three different minisatellite loci (D1S80, D17S30, ApoB), and 27 microsatellite loci of di-, tri-, or tetranucleotide repeats. Minisatellite instability was found in three (18%) tumors, with one of them exhibiting mutations in all three minisatellite loci, whereas two others showed mutations in one of two informative markers. By contrast, none of 20 sporadic thyroid cancers from patients with no history of radiation exposure was positive for minisatellite instability. Microsatellite analysis of post-Chernobyl tumors revealed a mutation in one (6%) tumor only at the locus of D10S1412, whereas all other 26 microsatellite markers showed identical patterns in each normal/tumor pair. Our results suggest that somatic cell microsatellite instability does not contribute to radiation-induced thyroid carcinogenesis. However, somatic minisatellite mutation events are present in a subset of radiation-induced, but not sporadic, thyroid cancers, suggesting that this type of genomic instability may play a role in radiation-induced tumorigenesis in the thyroid gland.

ret rearrangements in Japanese pediatric and adult papillary thyroid cancers.

Oncogenic rearrangements of the ret proto-oncogene (ret/PTC) are found uniquely in papillary thyroid carcinomas. The prevalence of ret/PTC in these tumors varies widely, from 0% to 87%, among patient series from different geographical regions. The differences in the prevalence of ret rearrangement have been ascribed to age, genetic, and/or environmental factors. The very high prevalence of ret/PTC in tumors arising in children after the Chernobyl nuclear accident has generated speculation that this oncogene may be an indicator of overt or inadvertent radiation exposure. In Japan, the prevalence of ret activation is reportedly quite low (0% to 9%). Here we examined the frequency of ret rearrangements in papillary carcinomas from Japanese adults and children by means of reverse transcription polymerase chain reaction (RT-PCR) followed by Southern hybridization. Ret rearrangements were detected in 4 of 11 (36%) tumors from the adult population, and in 3 of 10 (30%) pediatric tumors. One child with a solid variant papillary carcinoma had a ret-PTC3 rearrangement, further supporting the association between the solid variant histotype and this particular rearrangement of ret. The present data do not support a major geographic difference in the prevalence of ret/PTC rearrangements in papillary carcinomas between Japan, the United States, and Italy.

Overexpression of insulin-like growth factor-binding protein-4 (IGFBP-4) in smooth muscle cells of transgenic mice through a smooth muscle alpha-actin-IGFBP-4 fusion gene induces smooth muscle hypoplasia.

Insulin-like growth factor I (IGF-I) has been postulated to function as a smooth muscle cell (SMC) mitogen and to play a role in the pathogenesis of bladder hypertrophy, estrogen-induced uterine growth, and restenosis after arterial angioplasty. IGF-binding protein-4 (IGFBP-4) inhibits IGF-I action in vitro and is the most abundant IGFBP in the rodent arterial wall. To explore the function of this binding protein in vivo, transgenic mouse lines were developed harboring fusion genes consisting of a rat IGFBP-4 complementary DNA cloned downstream of either a -724 bp fragment of the mouse smooth muscle alpha-actin 5'-flanking region (SMP2-BP-4) or -1074 bp, 63 bp of 5'-untranslated region, and 2.5 kb of intron 1 of smooth muscle alpha-actin (SMP8-BP-4). SMP2-BP-4 mice expressed low levels of the exogenous IGFBP-4 messenger RNA (mRNA), which was not specifically targeted to SMC-rich tissue environments, and were therefore not analyzed further. Six SMP8-BP-4 transgenic lines derived from separate founders were characterized. Mating of hemizygous SMP8-BP-4 mice with controls produced about 50% transgenic offspring, with equal sex distribution. Expression of IGFBP-4 mRNA in nontransgenic littermates was maximal in liver and kidney. By contrast, transgenic IGFBP-4 mRNA expression, distinguished because of a smaller transcript size, was confined to SMC-containing tissues, with the following hierarchy: bladder > aorta > stomach = uterus. There was no transgene expression in skeletal muscle, brain, or cardiac myocytes. The abundance of IGFBP-4 measured by Western ligand blotting or by immunoblotting, was 8- to 10-fold higher in aorta and bladder of SMP8-BP-4 mice than in their nontransgenic littermates, with no change in plasma IGFBP-4 levels. Transgenic mice exhibited a significant reduction in wet weight of SMC-rich tissues, including bladder, intestine, aorta, uterus, and stomach, with no change in total body or carcass weight. In situ hybridization showed that transgene expression was targeted exclusively to the muscular layers of the arteries, veins, bladder, ureter, stomach, intestine, and uterus. Overexpression of IGFBP-4 was associated with SMC hypoplasia, a reciprocal phenotype to that of transgenic mice overexpressing IGF-I under control of the same promoter (SMP8-IGF-I). Double transgenic mice derived from mating SMP8-BP-4 with SMP8-IGF-I animals showed a modest decrease in wet weight at selected SMC tissues. Although we cannot exclude that the effects of IGFBP-4 may be IGF independent, these data suggest that IGFBP-4 is a functional antagonist of IGF-I action on SMC in vivo.

Targeted overexpression of IGF-I evokes distinct patterns of organ remodeling in smooth muscle cell tissue beds of transgenic mice.

Smooth muscle cells (SMC) of the vascular wall, bladder, myometrium, and gastrointestinal and respiratory tracts retain the ability to proliferate postnatally, which enables adaptive responses to injury, hormonal, or mechanical stimulation. SMC growth is regulated by a number of mesenchymal growth factors, including insulin-like growth factor I (IGF-I). To explore the function of IGF-I on SMC in vivo, the mouse SMC alpha-actin promoter fragment SMP8 (-1074 bp, 63 bp of 5'UT and 2.5 kb of intron 1) was cloned upstream of rat IGF-I cDNA, and the fusion gene microinjected to fertilized eggs of the FVB-N mouse strain. Mating of hemizygous mice with controls produced about 50% transgenic offspring, with equal sex distribution. Transgenic IGF-I mRNA expression was confined to SMC-containing tissues, with the following hierarchy: bladder > stomach > aorta = uterus > intestine. There was no transgene expression in skeletal muscle, heart, or liver. Radioimmunoassayable IGF-I content was increased by 3.5- to 4-fold in aorta, and by almost 10-fold in bladder of transgenic mice at 5 and 10 wk, with no change in plasma IGF-I levels. Wet weight of bladder, stomach, intestine, uterus, and aorta was selectively increased, with no change in total body or carcass weight of transgenic animals. In situ hybridization showed that transgene expression was exquisitely targeted to the smooth muscle layers of the arteries, veins, bladder, ureter, stomach, intestine, and uterus. Paracrine overproduction of IGF-I resulted in hyperplasia of the muscular layers of these tissues, manifesting in remarkably different phenotypes in the various SMC beds. Whereas the muscular layer of the bladder and stomach exhibited a concentric thickening, the SMC of the intestine and uterus grew in a longitudinal fashion, resulting in a marked lengthening of the small bowel and of the uterine horns. This report describes the first successful targeting of expression of any functional protein capable of modifying the phenotype of SMC in transgenic mice. IGF-I stimulates SMC hyperplasia, leading to distinct patterns of organ remodeling in the different tissue environments.

Distinct pattern of ret oncogene rearrangements in morphological variants of radiation-induced and sporadic thyroid papillary carcinomas in children.

In this study, we compare the morphological and genetic characteristics of 38 post-Chernobyl thyroid papillary carcinomas from Belarussian children 5-18 years old with those of 23 sporadic papillary carcinomas from the same age children without history of radiation exposure from Los Angeles and Cincinnati. Among radiation-induced tumors, solid variant of papillary carcinoma was found in 37%, follicular in 29%, typical papillary in 18%, and mixed and diffuse sclerosing variants in 8% each. In the sporadic group, a typical papillary pattern was prevalent in 70%, follicular in 17%, diffuse sclerosing variant in 9%, and solid in 4%. In both groups, the prevalence of ret rearrangements was high, but the frequency of specific types of rearrangement was significantly different. Among radiation-induced tumors, ret/PTC3 was found in 58%, ret/PTC1 in 16%, and ret/PTC2 in 3%, whereas among sporadic tumors, ret/PTC1 was found in 47% (P < 0.05), and ret/PTC3 was found in 18% (P = 0.01). The morphological variants of papillary carcinoma showed different prevalence of the specific types of ret rearrangement. Seventy-nine % of solid variant tumors had ret/PTC3, whereas only 7% had ret/PTC1 (P = 0.0007). Among typical papillary tumors, ret/PTC1 was found in 38%, ret/PTC3 in 19%, and ret/PTC2 in 5%. Thus, ret rearrangements are highly prevalent in pediatric papillary carcinomas from children exposed to radiation and in those occurring sporadically. However, the types of ret/PTC vary between these two populations, with ret/PTC3 present more commonly in post-Chernobyl tumors. Furthermore, solid variants have a high prevalence of ret/PTC3, whereas typical papillary carcinomas do not, suggesting that the different types of ret rearrangement confer neoplastic thyroid cells with distinct phenotypic properties.

Risk factors for thyroid cancer.

The potential risk factors for thyroid carcinoma development include genetic predisposition, exposure to therapeutic or environmental ionizing radiation, residence in areas of iodine deficiency or excess, history of preexisting benign thyroid disease, as well as hormonal and reproductive factors. In this review, we analyze some of the epidemiological data, as well as the possible molecular mechanisms by which certain environmental and genetic factors might predispose to thyroid tumorigenesis. (c) 1997, Elsevier Science Inc. (Trends Endocrinol Metab 1997; 8:20-25).