The clinical picture in this case unfolded across more than a decade, and the sequence of events illustrates why Felty syndrome remains one of the most diagnostically challenging entities in rheumatology and haematology. The patient’s rheumatoid factor was negative at 2 IU/mL, a detail worth emphasising because Felty syndrome has traditionally been associated with seropositive rheumatoid arthritis. Although anti-cyclic citrullinated peptide antibody testing was not documented, the seronegative presentation may have contributed to the initial diagnostic uncertainty, since clinicians encountering neutropenia and splenomegaly in a seronegative patient with inflammatory joint disease may consider a broader differential diagnosis, including lymphoproliferative disorders, before settling on Felty syndrome. The autoimmune haemolytic anaemia that marked her presentation six years before death added further complexity, as this cytopenia is not part of the classical triad and can point toward overlapping autoimmune and clonal lymphoproliferative processes.

The immunophenotypic findings in this case deserve careful consideration in light of what is known about gamma-delta T-cell biology. Gamma-delta T cells normally constitute a small fraction of the circulating T-cell pool, typically around one to five percent in peripheral blood, and they play important roles in epithelial surveillance and innate-like immune responses. Expansion of a gamma-delta population to approximately eleven percent of analysed cells in the initial splenic specimen, and to roughly twenty percent at splenectomy, represents a substantial deviation from normal proportions. The retrospective detection of this population in archival material is particularly instructive, because it demonstrates that flow cytometric data collected for other clinical purposes can yield critical diagnostic information when re-examined with a specific question in mind. Many laboratories archive flow cytometry data files, and systematic re-review of such datasets in patients with unexplained cytopenias and splenomegaly could potentially identify clonal T-cell populations earlier in the disease course.

The cytoplasmic granules observed in the atypical circulating lymphoid cells during the terminal phase reflect the cytotoxic nature of the neoplasm. Cytotoxic T cells contain specialised granules rich in proteins such as granzymes and perforin, and the expression of TIA-1, a granule-associated protein detected in both the splenic lymphocytes and the autopsy tumour cells, is a hallmark of this lineage. Cytotoxic T-cell large granular lymphocytic leukaemia, whether of alpha-beta or gamma-delta type, shares this phenotype, and the granular morphology visible on blood films is often one of the earliest morphological clues that a lymphocytosis is not merely reactive. In this patient, the emergence of conspicuous granules in atypical circulating cells coincided with the rapid rise in lactate dehydrogenase, a biochemical marker of high cellular turnover that ultimately exceeded 8000 U/L, roughly thirty-five times the upper limit of the reference range. Such extreme elevations of lactate dehydrogenase in lymphoid malignancies generally indicate high tumour burden, rapid cell turnover, or tissue destruction, and in this case served as a biochemical harbinger of the aggressive terminal phase.

The immunophenotype of the autopsy tumour cells, being positive for CD3, CD7, and TIA-1 but negative for CD5, CD4, and CD8, warrants comment. Loss of CD5, a pan-T-cell marker normally expressed on mature T cells, is a well-recognised aberrancy in T-cell large granular lymphocytic leukaemia and related clonal cytotoxic disorders. The double-negative CD4 and CD8 status, combined with the gamma-delta lineage suggested by the flow cytometric findings, is consistent with the double-negative gamma-delta cytotoxic phenotype that has been described in a subset of T-cell large granular lymphocytic leukaemia cases, particularly those associated with autoimmune conditions. The preservation of CD7, which is often retained in T-cell large granular lymphocytic leukaemia but frequently lost in more aggressive T-cell lymphomas, illustrates the difficulty of definitive subclassification in this terminal phase, a point the authors themselves acknowledged.

The molecular findings merit detailed discussion because they connect this case to a growing body of knowledge about somatic mutations in chronic clonal cytotoxic T-cell disorders. Activating mutations in STAT3, the gene encoding signal transducer and activator of transcription 3, are the most frequently identified somatic alterations in T-cell large granular lymphocytic leukaemia, detected in a substantial proportion of patients. The p.G618R substitution identified here lies within the SH2 domain, the region responsible for binding phosphorylated tyrosine residues on receptor tails. Mutations in this domain impair the negative regulatory mechanisms that normally keep STAT3 signalling transient, resulting in constitutive activation of downstream transcriptional programmes that promote survival and proliferation of the clonal cytotoxic cells. This sustained signalling helps explain the characteristic persistence and gradual expansion of the clonal population over years, and it also provides a mechanistic link to the autoimmune features that frequently accompany these disorders, since chronically activated cytotoxic T cells can contribute to neutropenia through antibody-independent mechanisms such as Fas-mediated apoptosis of neutrophil precursors.

The concurrent TET2 frameshift mutation adds another layer of biological interest. TET2 encodes a methylcytosine dioxygenase involved in DNA demethylation, and loss-of-function mutations in this gene are among the most common alterations in clonal haematopoiesis of indeterminate potential, an age-related phenomenon in which hematopoietic stem cell clones expand without overt malignancy. In the context of T-cell large granular lymphocytic leukaemia, TET2 mutations are frequent and may contribute to the survival advantage of the clonal cells, although their presence alone does not establish malignant transformation. The coexistence of a STAT3 mutation providing survival signalling with a TET2 mutation altering epigenetic regulation mirrors patterns seen in myeloid neoplasms, where the combination of a signalling pathway mutation and an epigenetic regulator mutation is a common theme in clonal evolution. The variant allele frequencies of 38.2 percent and 31.9 percent in the autopsy bone marrow indicate that the dominant neoplasm carried both mutations in a large fraction of cells, consistent with a clonal population in which these alterations were early founding events rather than late subclonal acquisitions.

The temporal relationship established by the allele-specific polymerase chain reaction analysis is arguably the most informative element of the molecular work. By demonstrating that STAT3 p.G618R was undetectable in the first three bone marrow examinations but detectable in the fourth and fifth, coinciding with the emergence of detectable T-cell receptor gamma-chain rearrangement, the investigators established a sequence in which the molecular clone became apparent only in the later phase of a disease that had been clinically stable for years. This stepwise pattern supports a model in which a small pre-existing clone, below the detection threshold of conventional testing, gradually expanded until it became demonstrable, and subsequently acquired additional genomic abnormalities, including the complex karyotype, that propelled the terminal aggressive transformation. The complex karyotype itself, with structural abnormalities of the X chromosome, duplications of chromosome 1 material, additions on chromosomes 2 and 5, and trisomy 9, reflects the accumulating genomic instability that often accompanies malignant progression in lymphoid neoplasms.

Several practical implications follow from this case for clinicians caring for patients with Felty syndrome or suspected T-cell large granular lymphocytic leukaemia. First, the absence of overt abnormalities on early bone marrow examinations does not exclude an incipient clonal process, and the retrospective identification of an expanded gamma-delta population in the initial splenic specimen suggests that careful multiparameter flow cytometry, with attention to gamma-delta T-cell proportions and aberrant marker expression, should be considered in patients with Felty syndrome who have unexplained or progressive cytopenias. Second, T-cell receptor gene rearrangement studies, which were not performed on the early bone marrow specimens in this case, can provide evidence of clonality that immunophenotyping alone may not establish, and their inclusion in the initial workup of suspected cases could shorten the diagnostic delay. Third, the detection of STAT3 mutations by sensitive methods such as allele-specific polymerase chain reaction can identify clonal disease before morphological or immunophenotypic evidence becomes apparent, although the presence of such mutations in patients without overt malignancy, as part of the spectrum of clonal cytosis of uncertain significance, means that positive results must be interpreted in clinical context.

The authors appropriately acknowledged limitations in their analysis, including the reliance on peripheral blood mononuclear cells as the comparator specimen because non-haematopoietic tissue was unavailable. This limitation is not trivial, since clonal haematopoiesis of indeterminate potential becomes increasingly common with advancing age, and low-level mutant clones in the blood could theoretically confound the interpretation of mutation testing in elderly patients. However, the complete absence of the STAT3 and TET2 mutations in the peripheral blood comparator, combined with their high variant allele frequencies in the splenectomy and autopsy specimens, strongly favours the interpretation that these mutations characterised the disease-associated clone rather than an incidental age-related process. The use of Sanger sequencing, which has a detection threshold of roughly ten to twenty percent variant allele frequency, means that smaller subclones carrying the same mutations would not have been detected, leaving open the possibility that the clone existed at low levels for longer than the molecular testing could demonstrate.

Beyond the individual case, this report contributes to a broader conceptual shift in how the boundary between autoimmune disease and clonal lymphoproliferation is understood. Felty syndrome and T-cell large granular lymphocytic leukaemia share not only clinical features, including neutropenia, splenomegaly, and association with rheumatoid arthritis, but also molecular features, with recurrent STAT3 mutations documented in both conditions. Some investigators have proposed that these entities represent points on a continuum rather than distinct diseases, with the proportion of clonal cells and the presence of additional genomic abnormalities determining where along that continuum a given patient falls. The present case, with its fourteen-year history of rheumatoid arthritis, its four-year clinical course diagnosed as Felty syndrome, and its terminal transformation into an aggressive clonal T-cell neoplasm, provides a longitudinal illustration of this continuum, showing how a chronic, indolent clonal process can remain clinically silent for years before acquiring the additional alterations that drive aggressive behaviour.

The rarity of aggressive transformation in this setting is itself noteworthy. Most patients with T-cell large granular lymphocytic leukaemia follow an indolent course, and the condition is generally managed conservatively, with treatment directed at the cytopenias rather than at eradication of the clone. Transformation to aggressive lymphoma is uncommon, and when it occurs, it is often associated with additional genetic events such as TP53 alterations, which were notably absent in this case despite the aggressive terminal course. This observation suggests that other genomic mechanisms, potentially involving the complex chromosomal abnormalities identified by cytogenetics, can drive aggressive progression independently of TP53, and it underscores the need for comprehensive genomic characterisation in such cases to identify the full complement of driver alterations. As larger cohorts of patients with Felty syndrome and associated clonal cytotoxic T-cell disorders are studied with modern sequencing technologies, the molecular determinants of transformation are likely to become clearer, potentially enabling risk stratification and earlier intervention in patients whose clonal populations show features associated with aggressive potential.

Subject of Research: Evolution of Felty syndrome-associated clonal cytotoxic T-cell disease into terminal aggressive T-cell lymphoma with shared STAT3 and TET2 mutations

Article Title: Felty Syndrome–Associated Clonal Cytotoxic T‐Cell Disease With Terminal Aggressive Progression: Shared STAT3 and TET2 Mutations Across a Prolonged Disease Course

Article References: Saito, K., Yokohama, A., Kwan Ee, O., Sorimachi, M., Imamura, K., Kobayashi, N., Miyazawa, Y., Ogawa, Y., Saito, T., Shirakura, T., Nobusawa, S., & Handa, H. (2026). Felty Syndrome–Associated Clonal Cytotoxic T‐Cell Disease With Terminal Aggressive Progression: Shared STAT3 and TET2 Mutations Across a Prolonged Disease Course. eJHaem, 7(5), Article e70406. https://doi.org/10.1002/jha2.70406

Image Credits: AI Generated

DOI: 10.1002/jha2.70406

Keywords: Felty syndrome, rheumatoid arthritis, T-cell large granular lymphocytic leukemia, STAT3 mutation, TET2 mutation, clonal hematopoiesis, TCR clonality, aggressive T-cell lymphoma, whole-exome sequencing, autopsy pathology, clonal evolution, neutropenia