T-cell large granular lymphocyte leukemia (T-LGLL) is a rare neoplasm, accounting for approximately 2-5% of chronic lymphoproliferative diseases in Western countries. It is characterized by clonal expansion of cytotoxic, often auto-immune reactive, mature T-cells (1). Next to T-LGLL, the 2016 World Health Organization classification of mature T- and Natural Killer (NK)-cell neoplasms, also lists the provisional entity of chronic lymphoproliferative disorders of NK cells (CLPD-NK) and aggressive NK-cell leukemia (ANKL) among monoclonal LGL proliferative diseases (2). T-LGLL, an expansion of CD3+ T-cell large granular lymphocytes (LGLs), is the most frequent variant representing ~85% of LGL proliferations and can further be subdivided into the common αβ-form and the rarer γδ-variant. Among the coreceptors CD8⁺ is usually more commonly expressed than CD4⁺. Additionally, mixed phenotype forms have been reported (3, 4). CLPD-NK accounts for approximately 10% and ANKL for approximately 5% of LGL proliferations. This case series excludes ANKL due to its clearly distinguished clinical and molecular features and different treatment approaches. We focus here on T-LGLL and CLPD-NK, both being referred to as LGLL.

The clinical presentation of LGLL is variable and typically includes cytopenias (particularly neutropenia and anemia), but often also symptoms of associated autoimmune disorders (mostly rheumatoid arthritis (RA), but also connective tissue diseases or vasculitis) (5). Furthermore, LGLL can be associated with secondary neoplasms, especially clonal B-cell expansions, but also solid cancers. The median patient age at diagnosis is 66 years, but approximately 15% of patients are younger than 50 years with an equal sex distribution (6, 7). Despite the course of LGLL being described as ‘indolent’, it is far from low-symptomatic and still associated with a shortened median overall survival (OS) of 9-10 years (8). Disease-related deaths are mainly due to severe infections. Such complications of the cytopenias and the autoimmune phenomena severely impair the quality of life of LGLL patients.

Diagnosis and management of LGLL is a challenge even for large academic centers. According to the WHO classification the diagnosis of LGLL requires a persistent (>6 months) increase in the number of peripheral blood (pB) LGL cells, usually 2-20 x 10⁹/L, without a clearly identified cause (2). Clonality is mandatory to be established and usually done by T-cell receptor (TCR) gene rearrangement studies (9, 10). In CLPD-NK monoclonality can indirectly be assessed by a restricted pattern of killer-cell immunoglobulin-like receptor (KIR) expression via flow-cytometric immunophenotyping, which is only done in few specialized laboratories. As a molecular hallmark, many LGLL harbor a genomic lesion of the signal transducer and activator of transcription 3 (STAT3). The gain-of-function STAT3 mutations D661 and Y640 account for two-thirds of such variants (9). Additionally, variants of STAT5B have been recognized in a minority of T-LGLL cases. Both mutations cause constitutive activation of the JAK/STAT signaling pathway (9, 11). While former studies did not find a clear impact of these lesions on clinical outcome (9, 11), a recent retrospective single-center analysis of a large LGLL cohort found an independent association of STAT3 mutations with shorter OS (4). More recently, missense mutations of the epigenetic regulator TET2 were identified as another major genomic hallmark in CLPD-NK (12, 13).

Overall, the diagnostic pitfalls in LGLL range from recognition of the subtle T-cell or NK-cell proliferation, repeated establishment of their clonality, distinction of the LGLL clone from normal (T-) lymphocytes by a unique immunophenotype as well as detection and interpretation of molecular aberrancies in the context of a commonly normal karyotype. Additional diagnostic challenges are imposed by a coexisting RA or by laboratory findings of an autoimmune hemolytic anemia (AIHA) or of a pure red cell aplasia (PRCA) or of myelodysplasia. Problems in differential diagnosis also expand to the differentiation from related conditions such as Felty-syndrome or from other mature T-cell leukemias/lymphomas such as T-cell prolymphocytic leukemia (T-PLL) or hepato-splenic T-cell lymphoma (HSTL) (14).

With respect to its therapeutic management, LGLL is considered incurable by currently available options, including immunosuppressive agents and low-dose chemotherapy. Treatment-defining prospective trials are hardly available. For a summary of tested strategies see (15). Furthermore, there is a great deal of uncertainty regarding the optimal timing of treatment initiation.

Here we present four challenging and instructive cases of LGLL that presented to our centers with typical as well as rare features of this heterogeneous disease. This case catalogue serves to emphasize numerous diagnostic pitfalls, unique clinical scenarios, and various therapeutic modalities. Typical characteristics and special features are presented in Table 1 .

Here we present heterogeneous presentations of T-LGLL and CLPD-NK that were seen in two academic institutions, with a focus on their diagnostic and therapeutic challenges. Our patients presented with unspecific symptoms, i.e. splenomegaly, autoimmune-mediated findings, or symptoms of cytopenias with a coincidental detection of LGL cells in flow cytometry.

In LGLL, often low-level lymphocyte infiltrations are misinterpreted as reactive, which frequently delays the definitive diagnosis. A thorough algorithm in the context of a fitting set of clinical presentations should include cytomorphology/histology, flow cytometry, a molecular clonality analysis, and gene-sequencing studies (11, 14, 16, 17).

The most common phenotype of T-LGLL is CD3⁺CD4^negCD5^+/lowCD8⁺CD16⁺CD57^dim. However, neither the described immunophenotype nor the morphological features of LGLs are entirely specific (14). Consequently, an accurate distinction from other mature T-cell disorders, e.g. early-phase (low proliferative) T-PLL, is highly relevant, prognostically and therapeutically, but can sometimes be difficult and requires the incorporation of diagnostic multi-parameter approaches. A prolymphocytic morphology is only found in ~60% of T-PLL and the post-thymic pan-T (CD2⁺CD3⁺CD5⁺CD7⁺) immunophenotype of T-PLL includes in a small fraction of cases also the T-LGLL-like CD4^-CD8⁺ pattern (14, 18). However, detection of a locus rearrangement involving a TCL1 gene (either TCL1A at chromosome 14 [mostly as an inv (14)] or MTCP1 at chromosome X) or proof of TCL1 protein expression in T-cells are established as unique major diagnostic criteria for T-PLL (14, 19–21)

CLPD-NK typically shows a CD3^-CD56⁺CD57^+/- immunoprofile. Cases of CD56^- CLPD-NK, as displayed in patients 1 and 3, have been described as well (22). CLPD-NK is associated with rather indolent courses and less often symptomatic than T-LGLL. Cytopenias and infections are characteristic as well as a higher incidence of second neoplastic diseases, however, the latter is observed across all subsets of LGLL.

There is a known association of LGLL with autoimmune conditions (23). Approximately one third of patients with LGLL suffer from rheumatoid arthritis (24) and less frequently from other autoimmune diseases like systemic lupus erythematosus, Sjögren syndrome, or autoimmune thyroid disorders (5). This association seems to be far less present in CLPD-NK (25). Consistent with the literature, patient 2 of our series had a history of chronic skin allergies and thyreoiditis while patient 4 suffered from ulcerative colitis.

Interestingly, one of our two CLPD-NK cases (patient 3) arose after an alloHSCT and the tumor cells were of donor origin. Self-limiting proliferations of LGLs after alloHSCT without clinical relevance have been described previously (26, 27), but aggressive forms of LGLL from donor cells were reported only in single case-reports (28–30) and in a series of four patients (31).

Of note, in both cases of CLPD-NK presented here, clonal TCR gene rearrangements were detected by PCR. This may seem contradictory, however, an analysis of KIR-restricted CLPD-NKs revealed TCR rearrangements in 50% of patients at first diagnosis (32).

We further presented a case of γδ T-LGLL (patient 2) and a case (patient 4) with a mixed phenotype of αβ/γδ LGLL. Interestingly, patient 4 developed the γδ clone during the course of the disease. In a recently published series of LGLL, the γδ variant accounted for approximately 15% of cases (4). Cases of αβ/γδ mixed phenotype are extremely rare and have so far been reported only as isolated cases or in very small series (3, 4). The diagnosis of γδ T-LGLL can be challenging; especially the differentiation from HSTL, which is often of γδ type and typically shows a low-level leukemic presentation. Particularly difficult are LGLL cases with absent or very low counts of clonal LGLL cells in pB and/or with a CD4-/CD8- phenotype (33). Detection of cytogenetic aberrations (isochromosome 7q or trisomy 8), described in >50% of HSTL (34), but atypical for T-LGLL, can be of assistance. HSTL affects predominantly younger men typically in the context of (medical) immunosuppression, often for a pre-existing autoimmune disease, yet the HSTL itself has not been associated with autoimmune phenomena. Nevertheless, the invariably aggressive course of HSTL sets it well apart from LGLL.

Another challenge is to define the cause(s) of cytopenias in LGLL, especially when these can also be caused by the therapeutic strategies. Patient 4 presented with anemia and underwent treatment with MTX and fludarabine. After becoming refractory to this treatment, further diagnostic workup was necessary to discriminate between therapy-related BM toxicity, AIHA, and PRCA, and eventually the diagnosis of PRCA was made. The association of LGLL with both AIHA (1, 17, 35) and PRCA (36) has been described and the accurate discrimination of those entities is necessary for the choice of an effective therapy (37, 38).

Although most LGLL patients do not require treatment at presentation, in 2/3 of cases therapy needs to be initiated at a later stage, especially due to severe neutropenias and subsequent infections, transfusion-dependent anemias, or thrombocytopenias (24). The current treatment options in LGLL are extremely limited. Standard approaches are based on supportive measures (e.g. transfusions, hematopoietic growth factors, antibiotics) and immunosuppressive therapies like MTX, cyclophosphamide, or cyclosporine (16) with limited evidence. The optimal sequence of MTX and cyclophosphamide is yet to be determined (ongoing trial NCT01976182).

Cyclophosphamide seems to show better efficacy in the control of symptoms and cytopenias as compared to MTX, but due to associated late toxicities it should not be administered for more than 12 months (39). Both agents need a minimum of 6-12 weeks before definite response assessments (17) and they both have treatment-associated cytopenias as side effects. In our case series, cyclophosphamide was administered in two patients and proved to be an active treatment option even for a living kidney donor. Reports on responses to other substances, like rituximab (40, 41) or the JAK1/3 inhibitors tofacitinib (42) and the JAK 2 inhibitor ruxolitinib (39) are sporadic and limited to small series. Generally, the treatment responses in LGLL are usually dissatisfactory being frequently incomplete and/or short-lived.

Another relevant aspect is the disease-inherent and treatment-related immunosuppression, which is of particular focus in current contexts of the COVID-19 pandemic. Although here not represented with a particular case, it has been repeatedly shown that patients with hematologic malignancies have a higher risk of severe or fatal COVID-19 infections (43–45). A single center retrospective analysis of 835 patients hospitalized with COVID-19, recently showed a significantly increased mortality of patients previously receiving immunosuppressive therapies (46). However, due to the rarity of LGLL, the exact morbidity and mortality risks related to COVID-19 infections in LGLL patients are unknown. Moreover, we do not know how disease and immunosuppressive therapies influence the effectivity of anti-COVID-19 vaccinations in LGLL patients. In other lymphatic malignancies, such as chronic lymphocytic leukemia (CLL), responses to vaccination are influenced by disease activity, current treatments, and previous therapies, especially regarding anti-CD20-antibodies (47, 48). A pragmatic strategy might be to adopt vaccination strategies from other hematologic malignancies like CLL, but not to neglect complex aspects of both humoral and cellular immunity specific for LGLL (49).

In summary, LGLL is a heterogenous group of diseases ranging from asymptomatic presentations over cases with severe impairments of quality of life, but long survival, to cases with significantly shortened life expectancy. In this series of selected cases with unique features, we illustrate pitfalls in the diagnosis, management, and treatment of LGLL of T-cell and NK-cell nature. In agreement with the literature, the uniqueness of the individual presentations and courses seems to override potential associations of clinical features, treatment responses, or outcomes with phenotype (αβ vs. γδ or T vs. NK) or genotype.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

Written informed consent was obtained from the individual(s) for the publication of any potentially identifiable images or data included in this article.

MH, NP, and VV contributed to conception and design of the study. AL, AS, DB, LW, NP, and VV acquired patient data and organized the database. NP and VV wrote the first draft of the manuscript. AL, AS, DB, and TB wrote sections of the manuscript. AD, AM, DJ, EB, MO, and TB provided diagnostic support. G-NF, MJ, SS, and UP provided administrative support. All authors contributed to the manuscript revision, read and approved the submitted manuscript.

The authors acknowledge support from the German Research Foundation (DFG) and Universität Leipzig within the program of Open Access Publishing” or “Dieser Artikel wurde durch den Publikationsfonds der Universität Leipzig und das Programm Open Access Publizieren der DFG gefördert”.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.