High frequency of STAT3 gene mutations in T‐cell receptor (TCR)γδ‐type T‐cell large granular lymphocytic leukaemia: implications for molecular diagnostics

T-cell receptor (TCR)γδ-type large granular lymphocytic leukaemia (LGLL) is a rare T-cell LGLL that is often complicated with cytopenia and autoimmune abnormalities.1-3 In CD8+ TCRαβ-type LGLL, a major type of T-cell LGLL, dysregulation of the Janus kinase/signal transducer and activator of transcription (JAK/STAT) signalling system with an interleukin 6 (IL-6)-dependent and/or independent manner, including activation of STAT molecules, has been recognised, and STAT3 is the most frequently mutated gene.4, 5 In CD4+ TCRαβ-type LGLL, STAT5B is frequently mutated.6 Mutations of the above-mentioned genes were located mainly in the src-homology (SH)-2 domain, and are considered to be activating mutations.7 On the other hand, the molecular basis of TCRγδ-type LGLL is still uncertain. We analysed the mutational profiles of the STAT3, STAT5B and tumour necrosis factor alpha-induced protein 3 (TNFAIP3) genes in TCRγδ-type LGLL. The present study was conducted in accordance with the Declaration of Helsinki and was approved by the Institutional Review Boards of Shinshu University School of Medicine. Written informed consent was obtained from the patients. Six patients with TCRγδ-type LGLL were recruited Table I. The median age was 48 years and five patients were male. The median haemoglobin value was 58 g/l. Neutropenia and pure red cell aplasia were observed in four cases each. The platelet count was within the reference range in all cases. The TCRγδ cells were positive for CD2 and CD3, and negative for CD4 and CD56. CD8 and CD16 positivity was detected in three cases each. The analysis of the TCRγ and TCRδ chain genes by the BioMed-2 protocol showed monoclonal patterns in all cases. Vγ9 and Vγ5 were utilised in two cases, Vγ2 in one case, Vδ1 in four cases and Vδ3 in two cases. Most of the TCRγδ T cells in the peripheral blood of normal subjects utilised a combination of the Vγ9 and Vδ2 chains, which were not found among the patient samples. We analysed the STAT3, STAT5B and TNFAIP3 coding regions of genomic DNA extracted from the mononuclear cells of the patients by deep amplicon sequencing as previously described.8 Briefly, amplicon sequencing was performed using Ion AmpliSeq technology on Ion PGM according to the standard protocol using the Ion 314 or 318 Chip Kit v2 (Thermo Fisher Scientific, Inc., Waltham, MA, USA). The data were analysed using the Torrent Suite software program (version 5.2.2). The main variant calling settings were as follows: variant frequency filter 0·005, base quality Q-value ≥20; minimum coverage of depth 1000; and maximum strand bias 0·95 (SNP), 0·9 (INDEL). Allele-specific polymerase chain reaction (PCR) for Y640F and D661Y of STAT3 and N642H and Y665F of STAT5B was performed as previously reported8 and the primers used are described in Table SI. The mean coverage of amplicon sequencing was 3,425x. All six cases were positive for STAT3 mutations Table I and no STAT5B mutations were recognised. The STAT3 mutation sites were H410R, Q448E, Y640F, D661Y and D661V Fig 1. The median (range) variant allele frequency (VAF) of STAT3 was 5·8 (1·3–22·6)%. In case 3, a STAT3 D661Y mutation was detected only by allele-specific PCR. Among the detected STAT3 mutations, four mutations (67%) were within the SH-2 domain. Case 3 was also positive for TNFAIP3 gene mutations at D117fs (VAF, 17·4%) and D119N (VAF, 17·3%). The other five patients were negative for TNFAIP3 mutations. These mutations were detected in TCRγδ T-cell fractions obtained by cell sorting in all four cases that were examined, but not in TCRαβ T cells. When compared with 50 other LGLL cases, which included 25 CD8+TCRαβ-type LGLLs, eight CD4+TCRαβ-type LGLLs and 17 chronic lymphoproliferative disorder of NK cells (CLPD-NK), the patients with TCRγδ-type LGLL were younger (P = 0·0498), had lower white blood cell counts and lymphocyte counts (P = 0·038, P = 0·024 respectively), and had lower haemoglobin values (P = 0·0033). The frequency of STAT3 mutations in TCRγδ-type LGLL was significantly higher than that in other LGLLs (6/6 vs. 22/50, P = 0·023) in an analysis using the system that was used in the present study Table SII. We detected STAT3 mutations in all the six patients by deep amplicon sequencing. In previous reports on TCRγδ-type LGLL, Sanger sequencing was applied for the analysis of the STAT3 gene, mainly the SH-2 domain, revealing that the rates of positivity for STAT3 mutations were between 25% and 67%.9, 10 The STAT3 mutated clones of TCRγδ-type LGLL were relatively small in size, with a median VAF of 5·8%, which implies that STAT3 mutation might occur as a late event in leukaemogenesis. However, the high incidence of STAT3 mutations in TCRγδ-type LGLL would suggest that the STAT3 mutations identified with high throughput sequencing could be diagnostic for and a biomarker of TCRγδ-type LGLL. In case 3, mutations in two genes, STAT3 and TNFAIP3, were recognised, which suggests that multiple mutational processes might play roles in the development of TCRγδ-type LGLL. Further analyses should be performed to clarify the entire mutational landscape of TCRγδ-type LGLL. In addition, TCRγδ-type lymphomas include hepatosplenic T-cell lymphoma, primary cutaneous γδT-cell lymphoma, and monomorphic epitheliotrophic intestinal T-cell lymphoma. Most patients with such lymphomas follow an aggressive clinical course, show resistance to various treatments, and have a poor prognosis. STAT5B is frequently mutated in these lymphomas,11, 12 in contrast to TCRγδ-type LGLL, which is usually clinically indolent. These different genetic features might be attributed to their distinct pathophysiology or clinical behaviour. In conclusion, STAT3 mutations occur with high frequency in TCRγδ-type LGLL, and their diagnostic implications should be confirmed in a larger cohort. The genetic-phenotypic relationship of the mutations in TCRγδ-type LGLL warrants further analyses. This research was supported in part by AMED under Grant Number JP18ek0109272. Fumihiro Ishida, Hideyuki Nakazawa and Taku Yamane designed the study and wrote the paper. Taku Yamane, Toru Kawakami, Jun Kobayashi, Fumihiro Kawakami and Yumiko Higuchi performed the experiments and analysed the data. Nodoka Sekiguchi, Toshimitsu Ueki, Hikaru Kobayashi, Sayaka Nishina, Hitoshi Sakai and Kazuo Oshimi collected the patients' samples and the data. All the authors read and approved the final version of the manuscript. Table SI. Primers for allele-specific PCR to detect STAT3 and STAT5B. Table SII. Comparison of TCRγδ-type T-cell large granular lymphocytic leukaemia with other types of large granular lymphocytic leukaemia. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.