HTT accounts for only 1% of all thyroid tumors with a female predominance and is most common in people in their 50 s [4, 7]. HTT is diagnosed based on histological findings of prominent trabecular or organoid growth patterns of elongated cells oriented perpendicular to the trabeculae with intratrabecular hyaline material [7, 8]. Neoplastic cellular proliferations are usually well-demarcated and exhibit nuclear features that overlap with PTC, such as nuclear grooves and intranuclear pseudoinclusions. In immunohistologic analysis, HTT exhibits membranous and cytoplasmic staining with MIB1, which can be diagnostic in differentiating HTT from PTC [9]. HTT is also positive for thyroglobulin and thyroid transcription factor-1 [7]. Focal and weak staining of HBME-1 and galectin-3 has also been reported. HTT is usually negative for calcitonin, CEA, chromogranin, and cytokeratins.

Cytologic diagnosis of HTT remains challenging, as it displays some cellular characteristics similar to PTC [10]. Although HTT exhibits several unique cytomorphological features, such as the presence of trabecular-like pattern, hyaline matrix, and intracytoplasmic yellow body, both HTT and PTC can exhibit hypercellularity, cellular atypia, intranuclear pseudoinclusions, and nuclear grooves on FNAC [7]. In the present case, in consideration that nuclear atypia was mild and nuclear groove was absent, the tumor was diagnosed as AUS on FNAC. In addition, the tumor was not diagnosed as HTT due to the lack of trabecular-like clusters, hyaline matrix, and yellow body. As we could not rule out the possibility of PTC with the results on FNAC and the ultrasound findings of the presence of microcalcification and paratracheal lymph nodes, hemithyroidectomy and central compartment dissection were performed.

In the present case, we used RT-PCR analysis and Sanger sequencing of a surgical specimen from a patient with HTT to clearly demonstrated the presence of a PAX8-GLIS3 fusion. PAX8, located on chromosome 2q14.1, is a paired box transcription factor that is highly expressed in differentiated thyroid follicular cells and required for normal thyroid development and function [11]. GLIS3, located on chromosome 9p24.2, belongs to the GLI-similar zinc finger transcription factor family and can function as either an activator or repressor of gene transcription. In the thyroid, GLIS3 is an important regulator of thyroid hormone biosynthesis [12]. The DNA-binding domain is encoded by exons 4–6 of GLIS3. The fusion point of the chimeric transcript of the PAX8-GLIS3 fusion was between exon 2 of the PAX8 gene and exon 3 of the GLIS3 gene. Therefore, the chimeric transcripts are regulated by the PAX8 gene promoter and preserve the zinc-finger-containing DNA-binding domains of GLIS3 [4]. The PAX8-GLIS3 fusion leads to significant overexpression of extracellular matrix-related genes (multiple collagen genes), which is likely responsible for excessive collagen synthesis and deposition in HTT [4]. The resulting considerable hyalinization gives these tumors their distinctive microscopic appearance and nomenclature.

To date, only 2 reports have demonstrated the presence of PAX8-GLIS3 fusions using surgical specimens from patients with HTT. Nikiforova et al. demonstrated for the first time in 2019 the PAX8-GLIS3 fusion in 93% of 14 patients with histologically confirmed HTT and the PAX8-GLIS1 fusion in 7% of these cases [4]. In the same year, a multi-institutional study confirmed the presence of the PAX8-GLIS3 fusion in 8 patients with histologically verified HTT; however, no thyroid tumors of varying histology, including PTC, harbored the PAX8-GLIS3 fusion [13]. Whole-exome sequencing, RNA sequencing, targeted NGS, fluorescence in situ hybridization, and immunohistochemistry have been used to detect PAX8-GLIS3 fusion [4, 13]. In the present case, we detected the PAX8-GLIS3 fusion using RT-PCR with Sanger sequencing confirmation, which is a more convenient approach than whole-exome and RNA sequencing. We also performed the NGS-based panel testing. However, the panel we used for NGS-based testing did not include the PAX8-GLIS3 fusion. We were able to exclude the presence of mutations in BRAF, RET, H/K/NRAS, and the TERT promoter as well as RET-PTC fusions, which are often detected by NGS in the thyroid tumors of varying histology. A previous report of one case described the detection of the PAX8-GLIS3 fusion by NGS genetic analysis of fine-needle aspirates from 180 thyroid nodules, leading to a diagnosis of HTT [14]. Molecular testing of FNAC samples for the markers including the PAX8-GLIS3 fusion can dramatically enhance the precision of preoperative FNAC.

The majority of HTT are clinically benign even after long-term follow-up, and complete excision is generally curative [2]. A study demonstrated no locoregional recurrence or distant metastasis in 29 cases with 6-year follow-up after surgery [15]. However, one case of 9 tumors with HTT was reported to have a neck lymph node metastasis [16]. In a study of 119 cases with a follow-up of 13–48 months, only one case with histologic findings of capsular and vascular invasion caused lung metastasis [17]. Therefore, we consider that long-term follow-up will be needed in the present case.

In conclusion, a rare case of HTT was demonstrated by imaging, cytologic, histologic, and molecular investigations. Molecular identification of the PAX8-GLIS3 fusion by RT-PCR with Sanger sequencing confirmation may aid in making the correct diagnosis.