Major Oncogenic Drivers and Their Clinicopathological Correlations in Sporadic Childhood Papillary Thyroid Carcinoma in Belarus

Abstract

Childhood papillary thyroid carcinoma (PTC) diagnosed after the Chernobyl accident in Belarus displayed a high frequency of gene rearrangements and low frequency of point mutations. Since 2001, only sporadic thyroid cancer occurs in children aged up to 14 years but its molecular characteristics have not been reported. Here, we determine the major oncogenic events in PTC from non-exposed Belarusian children and assess their clinicopathological correlations. Among the 34 tumors, 23 (67.6%) harbored one of the mutually exclusive oncogenes: 5 (14.7%) BRAF^V600E, 4 (11.8%) RET/PTC1, 6 (17.6%) RET/PTC3, 2 (5.9%) rare fusion genes, and 6 (17.6%) ETV6ex4/NTRK3. No mutations in codons 12, 13, and 61 of K-, N- and H-RAS, BRAF^K601E, or ETV6ex5/NTRK3 or AKAP9/BRAF were detected. Fusion genes were significantly more frequent than BRAF^V600E (p = 0.002). Clinicopathologically, RET/PTC3 was associated with solid growth pattern and higher tumor aggressiveness, BRAF^V600E and RET/PTC1 with classic papillary morphology and mild clinical phenotype, and ETV6ex4/NTRK3 with follicular-patterned PTC and reduced aggressiveness. The spectrum of driver mutations in sporadic childhood PTC in Belarus largely parallels that in Chernobyl PTC, yet the frequencies of some oncogenes may likely differ from those in the early-onset Chernobyl PTC; clinicopathological features correlate with the oncogene type.

Keywords: Chernobyl; clinicopathological characteristics; gene rearrangement; mutation; papillary thyroid carcinoma; sporadic childhood thyroid cancer.

Figure 1

Number of cases of thyroid cancer in children aged 0–14 years in Belarus during 1978–2014. Data are retrieved from the database of Childhood Cancer SubRegistry of Belarus. Approximate population of children 0–14 years old was 1.57 million in 1986.

Figure 2

Distribution of driver oncogenes by clinicopathological characteristics of 34 childhood PTCs. ¹ The dominant growth pattern: P—papillary, F—follicular, S—solid-trabecular.

Figure 3

Correspondence analysis of clinicopathological characteristics (black circles) of childhood PTCs with respect to oncogenic drivers (red arrows). The smaller angle between the types of driver mutation indicates their stronger correlation (e.g., between RET/PTC1 and BRAF^V600E), while right or obtuse angles indicate a lack of correlation (note the position of RET/PTC3 against other mutations). The smaller angle between clinicopathological variables indicates similarity in response pattern (e.g., an invasiveness score of 4 would be expected to frequently coexist with tumors displaying extrathyroidal extension). The distance between the positions of driver oncogenes and clinicopathological variables reflects, to some extent, the association between them (e.g., PTCs with RET/PTC3 are likely to display solid-trabecular, while those with RET/PTC1 display a papillary growth pattern more frequently). Dimension 1 and 2 accounted for 82.1% of variance cumulatively, and the dimensions 3–6 for the remaining 17.9% (not shown).

Figure 4

Characterization of the TBL1XR1/RET and TNIP1/RET fusions. (A) Genomic location, mRNA and protein structure, and validation by Sanger sequencing of TBL1XR1/RET. Exons 1–9 of TBL1XR1 are shown in blue and exons 12–20 of RET are shown in red for mRNA. For protein: LisH (lissencephaly type-1-like homology) domain, a dimerization motif; F-box-like, mediates protein–protein interactions; WD40, mediates protein–protein interactions; PTK, protein tyrosine kinase. (B) Genomic location, mRNA and protein structure, and validation by Sanger sequencing of TNIP1/RET. Exons 1–11 of TNIP1/RET are shown in blue and exons 12–20 of RET are shown in red for mRNA. For protein: CCD—coiled coil domain, mediates homodimerization; PTK, protein tyrosine kinase. Sequences of TBL1XR1/RET and TNIP1/RET are presented in Supplementary File S1 and are deposited to GenBank under accession numbers MZ269488 and MZ269489, respectively.

Figure 5

Tentative prevalence of BRAF^V600E (blue lines), RET/PTC1 (red lines), and RET/PTC3 (black lines) in radiation-related and sporadic childhood/pediatric/young PTC patients from Belarus and Ukraine from 1991 to 2009. Data are summarized from the current work and the sources are listed in Supplementary Tables S2 and S3. The length of lines tentatively corresponds to the periods of sampling.

Figure 6

Prevalence (%) of the major driver oncogenes analyzed in this work in childhood sporadic PTC from Belarus and those in other relevant groups of young patients with radiation-related (Chernobyl) and sporadic PTC. Percent mutant is presented on the same scale enabling direct visual comparison between different types of oncogenes. Each bar is labeled with corresponding prevalence. ND—not determined due to the lack of data. The most relevant/meaningful statistical comparisons are shown (all other comparisons returned statistically insignificant results).