Introduction

The telomerase reverse transcriptase (TERT) gene encodes the catalytic subunit of telomerase that is crucial to maintenance and regulation of the telomeres.1,2 In normal somatic adult tissues, telomerase activity is restricted to stem cells, and telomerase reactivation was proposed to be one of cancer hallmarks.3 Recently, hotspot somatic mutations in the promoter region of TERT, located −124 and −146 bp upstream from the ATG start site (c.−124C>T and c.−146C>T) were reported in several human cancers, including bladder (~85% of mutated cases), gliomas (~50%), thyroid (~15%) and melanoma (22–85%).4, 5, 6, 7, 8, 9 It was proposed that both c.−124C>T and c.−146C>T mutations create new binding motif sites (GGAA) of ETS transcription factors leading to upregulation of TERT levels and protein activity.4,5

Gastrointestinal stromal tumors (GISTs) are the most common mesenchymal tumor on the gastrointestinal tract characterized by hotspot mutations in KIT and PDGFRA genes, which are predictive of imatinib-based therapy response.10,11 Somatic BRAF mutations12, 13, 14 and germinative SDHx mutations were reported in a subset of KIT/PDGFRA wild-type GIST.15,16 Increased telomerase activity was reported in GISTs and was associated with poor prognosis.17 Yet, TERT promoter mutation has not been reported in GIST. Herein, we searched for the presence and clinicopathological association of the c.−124C>T and c.−146C>T TERT promoter mutations in a series of 130 bona fide GISTs.12,18, 19, 20

Materials and methods

One hundred and thirty cases of GIST were selected from the files of the Department of Pathology from Barretos Cancer Hospital, Brazil, Centro Hospitalar S. João and Garcia de Orta Hospital, Portugal. The cases were retrospectively re-evaluated and classified according to the WHO classification,21 and were assessed for the mean age, primary localization, tumor size, National Comprehensive Cancer Network (NCCN) risk classification,22 metastasis and overall survival. The mean age of the patients was 59.8 years, 52.3% were male and the tumors were located mainly in the stomach (50%) and the small intestine (32.7%). Most tumors had tumor size >5 cm, high malignancy risk and metastatic potential (Table 1).

Table 1 Association of TERT promoter mutation status and clinicopathological and molecular features of GISTs
Full size table

The characterization of the mutational status for KIT and PDGFRA was performed in all GISTs.12,18, 19, 20 In addition, the BRAF mutation status was evaluated in KIT/PDGFRA wild-type GISTs (n=9) from Barretos Cancer Hospital and the SDH genes status was evaluated in KIT/PDGFRA/BRAF wild-type GISTs (n=18) from Centro Hospitalar S. João.15,16

Tumor genomic DNA was extracted from formalin-fixed and paraffin-embedded tissues using the QIAamp DNA MicroKit (Qiagen, Hilden, Germany), following the manufacturer’s instructions.19 A fragment of the TERT promoter was amplified with PCR using primers 5′-AGTGGATTCGCGGGCACAGA-3′ and 5′-CAGCGCTGCCTGAAACTC-3′, resulting in a PCR product of 235 bp, which contained the chr5.hg19:g.1295228C>T and Chr5.hg19:g.1295250C>T sites of mutations. Alternatively, gene mutations can be designated based on their upstream location to the ATG initiation codon of TERT, as c.−124C>T, and c.−146C>T, as previously described.7 PCR was performed with an initial denaturation at 95 °C for 15 min, followed by 40 cycles of 95 °C denaturation for 30 s, 64 °C annealing for 90 s and 72 °C elongation for 30 s and 72 °C final elongation for 7 min. Quality of PCR products was confirmed with gel electrophoresis. DNA sequencing of the PCR product was performed using the BigDye Terminator version 3.1 cycle sequencing kit (Applied Biosystems, Foster City, CA, USA) and an ABI PRISM 3500 xL Genetic Analyzer (Applied Biosystems). The chromatograms were compared with the reference sequence (GeneBank, TERT: ENST00000310581). The SPSS 19.0 software (IBM Corp, Armonk, NY, USA) was used for all statistical analysis. To assess the relationship between variables, we used the Fisher’s exact test. The P-value established for the statistics significance was <0.05.

Results

We found TERT promoter mutations in 3.8% (5/130) of the GISTs (Table 2 and Figure 1). The identified mutations are described in the LOVD database (https://research.cchmc.org/LOVD2/home.php;patient IDs 819–823). The c.−124C>T mutation was the most common event, present in 2.3% (3/130), and the c.−146C>T mutation in 1.5% (2/130) of GISTs. The two mutations occur in a mutually exclusive manner. No statistical correlation was found between TERT mutation and GIST clinical or molecular features (Table 1). Yet, TERT mutations appeared in tumors of slightly older patients, and no TERT-mutated cases were detected in benign/very-low malignancy risk GISTs (Table 1).

Table 2 Clinicopathological and molecular data of the GISTs with TERT promoter mutation (clinicopathological and molecular features of TERT promoter-mutated GISTs)
Full size table
Figure 1
figure 1

Electropherogram of TERT promoter mutations. (a) Heterozigotic c.−124C>T mutation (arrow); (b) Heterozigotic c.−146C>T mutation (arrow).

Full size image

Discussion

This study describes for the first time the occurrence of TERT promoter mutations (c.−124C>T and c.−146C>T) in GISTs, being present in ~4% of the cases. As paired blood or constitutive DNA of the tumors analyzed in the present study was not available, we cannot confirm the somatic nature of the c.−124 or c.−146 mutations identified. However, germline mutations at these hotspots were not reported in the various TERT studies that performed such paired (tumor versus normal) analysis.4, 5, 6, 7,23, 24, 25 In addition, in the COSMIC database,26 these mutations are described as somatic, and they are not present in the 1000 Genomes database.27 Therefore, we can almost certainty assume that the mutations observed in GISTs were somatically acquired.

Previously, we analyzed a series of 36 GISTs and did not identify any TERT promoter mutation.7 Likewise, Killela et al6 also analyzed nine GISTs and did not found any TERT promoter mutation. As identical methodologies were used in all studies, one plausible reason for this discrepancy is the small number of cases analyzed in the previous studies.6,7

TERT promoter mutations seem to be widespread in cancer, although showing tissue specificity. Killela et al6 suggested that cancers developing in tissues that are regularly self-renewing, such as in the gastrointestinal tract, skin and bone marrow, are not likely to harbor telomere-maintaining mutations, as telomerase is already epigenetically activated in their precursor cells. In contrast, cancers arising from cells that are not regularly self-renewing might harbor such mutations.6 GISTs fit to the second setting, as they are assumed to originate from the low-renewal Cajal cells or their precursors.28 GISTs are prone to exhibit a high risk of disease relapse and metastasis spreading to distant organs such as the liver, peritoneal surface and lung.10 Previous reports associate telomerase activity in GIST with higher tumor malignancy risk, metastasis and worse prognosis.29, 30, 31 We found a low frequency of TERT mutations in GIST, but any statistical association was found with tumor aggressiveness; however, most TERT-mutated GISTs displayed high recurrence risk features. Although our series is undersized to allow definitive conclusions, it would be of interest to further evaluate whether TERT promoter mutations associate with a higher expression of telomerase in GISTs, and to assess whether TERT promoter mutations associate with poor prognosis as reported in other cancers such as cancers of the thyroid,32 melanoma9 and brain.6

On the whole, our study establishes the presence of TERT promoter mutations in a subset of GISTs (4%). Future studies are required to validate our findings and to elucidate the potential biological and clinical impact of TERT promoter mutation in GIST pathogenesis.