The Strongest Genetic Signal for Thyroid Cancer — FOXE1 and the 9q22 Regulatory Hub
Among the hundreds of genetic variants studied for thyroid disease, rs965513 at chromosome 9q22.33 stands alone as the strongest common-variant association with thyroid cancer risk ever identified | Gudmundsson et al. Nature Genetics 2009; OR 1.75, P=1.7x10-27. But this variant does more than influence cancer: it also shifts thyroid hormone levels in the general population, lowering TSH and T4 while raising T3. The locus sits within PTCSC2 | Papillary Thyroid Cancer Susceptibility Candidate 2, a long noncoding RNA gene whose transcripts are downregulated in thyroid tumors, a long noncoding RNA gene, approximately 57 kb upstream of FOXE1 | Forkhead Box E1, also known as thyroid transcription factor 2 (TTF-2), essential for thyroid gland development and differentiation — a transcription factor indispensable for thyroid gland formation during embryonic development.
The Mechanism
FOXE1 (also called TTF-2) is one of a small number of transcription factors that orchestrate thyroid organogenesis. FOXE1 knockout mice lack a thyroid gland entirely | Demonstrating the gene's essential role in thyroid morphogenesis and develop severe hypothyroidism. In adults, FOXE1 maintains the differentiated state of thyroid follicular cells.
The rs965513 variant resides within an intron of the PTCSC2 lncRNA gene, in a linkage disequilibrium block containing multiple enhancer elements | He et al. PNAS 2015 identified at least three regulatory elements overlapping the FOXE1 promoter that regulate FOXE1 expression. The risk A allele is associated with reduced expression of three genes simultaneously | PTCSC2 (P=0.036), FOXE1 (P=0.012), and TSHR (P=0.024) in unaffected thyroid tissue: PTCSC2, FOXE1, and — notably — TSHR (the TSH receptor). The molecular mechanism involves MYH9 (myosin-9) binding to PTCSC2 | He et al. PNAS 2017 showed MYH9 suppresses the bidirectional FOXE1/PTCSC2 promoter, while PTCSC2 reverses this inhibition at the shared bidirectional promoter of FOXE1 and PTCSC2. When PTCSC2 expression is reduced by the risk allele, MYH9-mediated suppression of FOXE1 goes unchecked, further reducing thyroid-protective gene activity.
The downstream consequence is twofold: reduced FOXE1 impairs thyroid cell differentiation and interferes with the p53 tumor suppressor pathway | Affecting IGFBP3 and THBS1 expression in primary thyroid cells, while reduced TSHR expression alters thyroid sensitivity to TSH stimulation, explaining the variant's effects on circulating thyroid hormone levels.
The Evidence
The landmark Gudmundsson et al. 2009 GWAS | Gudmundsson et al. Common variants on 9q22.33 and 14q13.3 predispose to thyroid cancer in European populations. Nature Genetics 2009 identified rs965513 with an odds ratio of 1.75 for differentiated thyroid cancer (P=1.7x10-27) in 192 Icelandic cases with replication in European cohorts. The same study found that each A allele was associated with a 5.9% decrease in TSH, a 1.2% increase in T3, and a 1.2% decrease in T4 in the general population.
A comprehensive meta-analysis of 23 studies | Wang et al. Oncotarget 2016, covering 20,736 cases and 142,400 controls across 13 countries confirmed the per-allele OR at 1.58 (95% CI 1.32-1.90), with Caucasians showing higher risk (OR 1.65) than Asians (OR 1.49). Under the recessive model, AA homozygotes had an OR of 2.10 (95% CI 1.66-2.64) compared to G-allele carriers, and under the homozygous model, AA vs GG showed an OR of 2.80 (95% CI 2.12-3.69).
A European cohort study of 277 cases and 309 controls | Penna-Martinez et al. Endocrine Oncology 2021 replicated the finding with a per-allele OR of 1.61 (95% CI 1.27-2.04), showing a clear allele-dosage effect: heterozygotes had OR 1.66 and AA homozygotes had OR 2.93 compared to GG.
The variant was also identified as a major genetic determinant for radiation-related thyroid cancer | Takahashi et al. found OR 1.65 for radiation-associated papillary thyroid cancer in Chernobyl survivors in Chernobyl survivors, suggesting that genetic background at this locus modifies radiation-induced thyroid cancer risk.
Regarding thyroid function in the general population, large GWAS studies have confirmed rs965513 as a significant locus for both TSH and free T4 levels. Individuals with the AA genotype have measurably lower TSH than GG carriers, consistent with reduced TSHR expression shifting the thyroid set point.
Practical Implications
The actionability of this variant differs depending on context. For thyroid cancer risk, the absolute risk increase is modest — thyroid cancer is relatively uncommon (annual incidence ~14 per 100,000), so even a 2-3 fold relative increase in AA carriers translates to a small absolute risk. However, in individuals with other risk factors — family history of thyroid cancer, history of radiation exposure, or thyroid nodules on imaging — this genotype can inform surveillance decisions.
For thyroid function, the variant's effect on TSH set point means that A-allele carriers may have naturally lower TSH levels. This is important context when interpreting thyroid function tests: a TSH in the low-normal range may be constitutionally appropriate for these individuals rather than a sign of subclinical hyperthyroidism.
Interactions
The rs965513 locus interacts with other thyroid-related variants through the shared regulatory architecture. The variant's simultaneous effect on FOXE1, PTCSC2, and TSHR expression means it influences both thyroid development/cancer pathways and thyroid hormone regulation. Individuals carrying both rs965513-AA and DIO2 rs225014-CC (Thr92Ala) may experience compound effects on thyroid function: reduced TSHR sensitivity from rs965513 combined with impaired T4-to-T3 conversion from DIO2 could create a more pronounced mismatch between standard thyroid function tests and tissue-level thyroid hormone activity. This interaction has not been formally studied but is biologically plausible given the converging pathways.
The Gudmundsson 2009 study found that individuals homozygous for both rs965513 and the 14q13.3 thyroid cancer variant had a 5.7-fold greater thyroid cancer risk compared to non-carriers at either locus, demonstrating multiplicative effects between thyroid susceptibility loci.
All Genotypes
No increased thyroid cancer risk from this locus — standard thyroid function expected
You have two copies of the G allele, which is the common protective allele at this locus. This genotype is found in approximately 51% of people of European descent and over 86% of East Asians. You do not carry the variant associated with increased thyroid cancer susceptibility or altered TSH set point at this locus. Your FOXE1, PTCSC2, and TSHR expression in thyroid tissue is expected to be at normal levels.
One copy of the risk allele — moderately increased thyroid cancer susceptibility
You carry one copy of the A risk allele and one copy of the protective G allele. This is the most common genotype among people of European descent, found in approximately 40% of the population. Heterozygotes have a per-allele OR of approximately 1.58-1.66 for differentiated thyroid cancer compared to GG homozygotes. Your TSH set point may be slightly lower than average due to the A allele's effect on TSHR expression in thyroid tissue, though the shift is smaller than in AA carriers.
Two copies of the risk allele — approximately 2-3 fold increased thyroid cancer risk and altered TSH set point
You have two copies of the A allele at the strongest GWAS thyroid locus ever identified. This genotype is found in approximately 9-11% of people of European descent, and less than 1% of East Asians. Multiple meta-analyses consistently show that AA carriers have approximately 2.1-2.9 fold increased risk of differentiated thyroid cancer (papillary and follicular types) compared to GG carriers. This variant also affects your thyroid hormone set point: each A allele is associated with lower TSH and T4 levels and slightly higher T3, meaning your naturally lower TSH should not be mistaken for subclinical hyperthyroidism.