Restricted Expression of the Constant Region 1 of T‐Cell Receptor β by Flow Cytometry Facilitates Detection of T‐Cell Neoplasms With High Specificity but Moderate Predictive Value

ABSTRACT

Objectives

To determine the diagnostic value of disclosing monotypic T‐cell populations by expression analysis of the constant region 1 of T‐cell receptor β (TRBC1) by flow cytometry, for the detection of T‐cell neoplasms, in a routine hematopathology practice setting.

Methods

A panel of antibodies ECD‐CD16/PECy5.5‐CD4/PECy7‐CD2/APCAF700‐CD7/KO‐CD8/FITC‐TCRγδ/PE‐TRBC1/APC‐CD26/APC‐CD57/APC‐CD30/APCH7‐CD45/BV421‐CD3 was applied. One thousand and twenty‐nine cases investigated during 26 months were reviewed and categorized as monotypic or polytypic depending on the expression of TRBC1 in all identifiable CD3+ T‐cell subsets.

Results

Although TRBC1 restriction proved to be sensitive (90.7%) with a high negative predictive value (99.1%), in some T‐cell neoplasms, particularly those with significant inflammation, including angioimmunoblastic T‐cell lymphoma, monotypic T‐cells were not detected. The specificity was high (95.6%) albeit with a modest positive predictive value (67.1%), reflecting T‐cell clones of uncertain significance (T‐CUS) without T‐cell neoplasm. Of these clones, 53.6% exhibited a T‐cell large granular lymphocytic leukemia‐like phenotype, most commonly found in the peripheral blood or bone marrow, while others were identified in association with unrelated primary tumors or other comorbidities. Seventy‐nine percent of the T‐CUS cases remained stable over time during 1–77 months of follow‐up.

Conclusion

Including TRBC1 antibodies in a routine flow cytometry panel facilitates the identification of T‐cell neoplasms. The analysis must be interpreted within its clinical context since T‐cell lymphomas with a small and sometimes surface CD3‐negative neoplastic T cell population, may display normal patterns of TRBC1‐expression with a background of reactive T cells. Conversely, monotypic T‐cells can be found in the absence of T‐cell neoplasm.

1. Introduction

T‐cell neoplasms are typically aggressive tumors that may be challenging to diagnose due to overlapping morphologic and immunophenotypic features with reactive (inflammatory) lymphoproliferations. Therefore, clonality analyses play an important role in establishing a diagnosis of T‐cell lymphoma. Most T‐cell lymphomas are of αβ‐origin and the most common nodal subtypes include peripheral T‐cell lymphoma not otherwise specified (PTCL NOS), angioimmunoblastic T‐cell lymphoma (AITL) and anaplastic large cell lymphomas (ALCL) that are either ALK‐positive or ALK‐negative [1]. These T‐cell lymphomas typically involve lymph nodes, but manifestations can also be found at extranodal sites as well as in peripheral blood (PB), bone marrow (BM), and body cavity fluids. Other T‐cell neoplasms, such as T‐cell large granular lymphocyte leukemia (T‐LGLL), preferentially involve PB and BM. Blood manifestations can also be seen in patients with cutaneous T‐cell lymphomas including mycosis fungoides (MF). Polymerase chain reaction (PCR), today replaced by next‐generation sequencing (NGS) at many institutions, was for a long time considered the gold standard for determination of T‐cell receptor gene rearrangements in everyday hematopathology practice [2]. Depending on the diagnostic context, such as the quality of available histological material, including BM and lymph node specimens, clonality analysis by PCR may or may not be required to exclude or confirm a diagnosis of T‐cell lymphoma. Previously available corresponding immunophenotypic assays targeting the TCR‐Vbeta chain with multiparameter flow cytometry (MFC) were quite cumbersome [3, 4]. However, the application of an antibody specific for one of two mutually exclusive T‐cell receptor β‐chain constant domains, that is, TRBC1, provided an opportunity to detect monoclonal (monotypic) αβ T‐cells also by MFC [5]. Nonneoplastic αβ T cells comprise a mixture of TRBC1‐ and TRBC2‐expressing cells, much like kappa‐ and lambda‐expressing nonneoplastic B‐cells, whereas αβ T‐cell lymphomas are expected to consist of cells that are monoclonal for one β chain constant region variant. From a diagnostic perspective, it is therefore highly relevant to investigate if immunophenotypic clonality analysis based on the expression of TRBC1 is feasible and comparable to established PCR methods. Several recent reports have demonstrated the excellent ability of TRBC1 restriction analysis to confirm or exclude T‐cell neoplasms [6, 7, 8, 9, 10, 11, 12, 13].

To determine the accuracy of TRBC1 restriction analysis by MFC in real‐life hematopathology practice, we reviewed 1229 consecutive routine cases investigated with a panel including the commercially available JOVI‐1 TRBC1 antibody during a period of 26 months. Our results provide an evaluation of the utility of TRBC1 restriction analysis in the routine diagnostic work‐up of T‐cell neoplasms and highlight some clinically relevant limitations.

2. Materials and Methods

2.1. Patients

All clinical samples investigated using the MFC panel including the TRBC1 from November 2019 to December 2021 at the Department of Clinical Pathology in Lund were included. Rare γδ T‐cell lymphomas were excluded because of the absence of β‐chains. The MFC panel including the TRBC1 antibody was applied when there was a clinical suspicion of T‐cell lymphoma, when an aberrant T cell population was detected in the screening panel [14, 15], when the CD4/CD8 ratio was (> 10 or < 0.5) and in some BM samples analyzed for pancytopenia to rule out T‐cell lymphoma.

The T‐cell lymphoma diagnosis was established by a hematopathologist using WHO 2017 criteria combining morphological, immunophenotypical, molecular, and clinical data [16]. Ethical approval for the study was granted by the Swedish Ethical Review Board (No. 22‐02948‐01).

2.2. Flow Cytometry

PB and BM samples were received in heparin tubes and fine needle aspirations (FNA) in phosphate‐buffered saline. Small pieces of lymph node tissue were ground to produce a cell suspension. FNA, body fluids and cell suspensions were supplemented with TransFix [17]. All samples were processed within 72 h after collection.

White blood cells from PB, BM aspirates, fine needle aspirates or fresh tissue were counted on a Sysmex XN‐350 hematology analyzer and diluted to a white blood cell count of less than 15 × 10⁶/mL. Red blood cells from PB and BM specimens were bulk‐lysed using Pharm‐Lyse (BD) according to the manufacturer's recommendations.

First, samples were analyzed with 10‐color 14 or 15 antibody screening panels as published previously [14, 15]. Next, aliquots of 100 μL were incubated at room temperature in the dark for 10 min with a cocktail containing 10 monoclonal antibodies: ECD‐CD16, PECy5.5‐CD4, PECy7‐CD2, APCAF700‐CD7, KO‐CD8 (Beckman Coulter, Miami, FL, USA), FITC‐TCRγδ, APC‐CD26 or CD57 or CD30, APCH7‐CD45, BV421‐CD3 (BD Biosciences, San Jose, CA, USA) and PE‐TRBC1 (Biolegend, San Diego, CA, USA). CD26 was included (at the expense of CD57 and CD30) to analyze Sézary clones in the PB from patients with MF or Sézary syndrome. CD30 was chosen whenever there was a suspicion of Hodgkin lymphoma, PTCL NOS or ALCL. For the remaining analyses, CD57 was chosen to disclose T‐LGLL‐like clones. When an atypical phenotype (e.g., CD3−/CD5+ or CD3+/CD5−) was discovered in the screening tube, CD5 was included in the extended T‐cell panel instead of CD26, CD57, or CD30.

Stained cells were resuspended in 200 μL of PBS, and at least 100 000 events were acquired on Navios flow cytometers (Beckman Coulter, Miami, FL, USA) within 60 min. Instrument settings on the three flow cytometers used at the laboratory were harmonized according to the Harmonemia protocol [18] to ensure reproducibility of data.

Listmode files were analyzed on Kaluza version 2.1 (Beckman Coulter, Brea, CA, USA). Briefly, sequential manual gating with color coding was used to identify each major T cell population, based on the most biologically relevant and informative parameters, that is, CD2, CD3, CD4, CD7, CD8, CD26, CD57, CD30 and TCR γ. TCR γδ populations were excluded from the analysis by appropriate gating. Discrete T‐cell subsets were visually identified and manually gated. Only distinct clusters of T‐cell subsets comprising 40 or more events were studied. TRBC1 expression was examined in detail on subsets of CD3+, CD4+, CD8+, CD4+/CD8+, CD2+, CD7+, and either of CD57+, CD26+, CD30+, or CD5+ cells. Restricted (monotypic) TRBC1 expression was defined as the presence of TRBC1 on > 85% or < 15% of cells within a distinct T‐cell population or, in rare cases, as the presence of a dominant TRBC1‐dim cluster [6, 12, 13, 19]. The 15%/85% threshold is commonly used although another study determined slightly different cut‐offs for the detection of TRBC1 restricted populations [20]. A typical gating strategy to disclose discrete T‐cell populations with TRBC1 restriction analysis is shown in Figure S1.

Our TRBC1 protocol was developed just prior to this study and essentially adheres to current recommendations regarding fluorochrome selection, antibody optimization and interpretation including the 15%/85% threshold for detecting a clonal TRBC1‐restricted population [19].

2.3. T‐Cell Receptor Gene Rearrangement Analysis

PCR for TCR gene rearrangements was performed on selected cases based on the hematopathologist's assessment, if it was deemed necessary for the final diagnosis. Cellular DNA was extracted with subsequent PCR amplification performed in five multiplex PCR tubes according to the BIOMED‐2 protocol [21] with primers targeting the TCR Vγ, Jγ, Vβ, Dβ, and Jβ regions (Invivoscribe Technologies, San Diego, CA). The amplified fragments were separated and detected by capillary gel electrophoresis (ABI Prism 3130 Bioanalyzer, Applied Biosystems, Warrington, UK).

2.4. Statistical Analysis

Sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) of the TRBC1 restriction analysis were calculated [22, 23]. Pearson chi‐square test was applied to assess immunophenotypical differences between distinct T‐cell populations [24]. The software program used was IBM SPSS statistics 29.0 (Chicago, IL).

3. Results

3.1. Predictive Value of TRBC1 Analysis

Of 1229 samples analyzed with the T‐cell panel, 127 fulfilled criteria for diagnosis of T‐cell lymphoma (Tables 1 and S1). Of these, 98 cases were surface CD3+ and showed monotypic TRBC restriction while 10 were polytypic (Table S1) with the TRBC1 assay (discussed in detail below). The remaining 19 cases were surface CD3‐negative T‐cell neoplasms. Thus, the sensitivity of monotypic TRBC restriction with respect to bona fide T‐cell neoplasms was 90.7% or 92.1% if also surface CD3‐negative cases were included in the analysis (Table 1). An example of a typical TRBC1 restricted T‐cell lymphoma (PB manifestation of CD30+ ALK‐negative ALCL) is shown in Figure 1. However, of the 146 cases showing monotypic TRBC restriction in the study (79 cases with and 67 cases without TRBC1 expression, Table S1), 48 (56 with surface CD3‐negative cases included) did not fulfill T‐cell neoplasm criteria, reflecting the modest PPV of the method—67.1% or 67.6% with surface CD3‐negative cases included (Table 1). Hence, about one third of the surface CD3‐positive cases (33%) with TRBC restriction were non‐neoplastic (Table 1). When cytology samples (180 FNA and eight body fluids including four cerebrospinal fluid samples, and four pleural effusions) were analyzed separately, sensitivity reached 85.7% or 90.0% when surface CD3‐negative cases were included. PPV was 70.6% or 75.0% when including the surface CD3‐negative cases (Table S1). All four cerebrospinal fluid samples were malignant, whereas only one of the pleural effusions was malignant.

TABLE 1.

Diagnostic performance of TRBC1 restriction analysis by flow cytometry in 1229 consecutive cases with or without T‐cell neoplasm (T‐cell lymphoma or T‐LGL leukemia excluding gamma/delta T‐cell neoplasms), excluding or including surface CD3‐negative (sCD3−) cases.

	Lack of TRBC restriction (polytypic), n = 1056	TRBC restriction (monotypic) excluding sCD3− cases, n = 146	TRBC restriction including sCD3− cases, n = 173
No T‐cell neoplasm, n = 1102	1046	48	56
T‐cell neoplasm, n = 127	10	98	117
Sensitivity (%)		90.7	92.1
Specificity (%)		95.6	94.9
False‐negative rate (%)		9.3	7.9
False‐positive rate (%)		4.4	5.1
NPV (%)		99.1	99.1
PPV (%)		67.1	67.6

FIGURE 1.

Patient with ALK‐negative anaplastic large cell lymphoma in blood displaying monotypic T‐cells. (a) Monotypic expression of TRBC1 on CD30+ lymphoma cells by flow cytometry. (b) Monoclonal TCR β gene rearrangement by PCR. (c) CD30 staining on bone marrow clot. CD, cluster of differentiation; PCR, polymerase chain reaction; TCR, T‐cell receptor; TRBC1, constant region 1 of T‐cell receptor β.

One thousand and forty‐six cases (85.1%) were polytypic (polyclonal) as determined by the expression pattern of TRBC1 and negative for T‐cell lymphoma, confirming the high negative predictive value (NPV) of a polytypic TRBC1 result with respect to T‐cell lymphoma diagnosis (Table 1).

In a minority of the polytypic cases (35; 3.3%), clinical and/or morphological features raised suspicion of a possible T‐cell neoplasm. For these reasons, PCR‐based clonality analysis was deemed necessary to exclude T‐cell malignancy with 28/35 PCR analyses confirming polyclonal T‐cell population in these samples (Table 2). Nevertheless, seven of the 35 samples displayed monoclonal T‐cell populations as judged by PCR (Table 2) despite the polytypic pattern with the TRCB1 antibody, of which four were finally diagnosed as T‐cell neoplasms (see below).

TABLE 2.

TRBC restriction by flow cytometry vs. PCR results in all cases with both PCR and flow cytometry available (n = 114). NPA, negative percent agreement; PPA, positive percent agreement.

	Lack of TRBC restriction (polytypic), n = 35	TRBC restriction (monotypic), n = 79
PCR polyclonal, n = 46	28	18
PCR monoclonal, n = 68	7	61
NPA (%)	60.1
PPA (%)		89.7

3.2. Polytypic Expression of TRBC1 Does Not Exclude T‐Cell Neoplasm

The 10 cases of T‐cell lymphoma where TRBC1 restriction could not be demonstrated by MFC were reanalyzed in detail. For six of these, the discrepancy could be explained by either a paucity of lymphoma cells or concurrent fibrosis making MFC detection difficult or by absence of CD30 in the panel for identification of ALCL cells (Cases 1–6, Table 3).

TABLE 3.

T‐cell neoplasms without detectable TRBC‐restricted subsets by flow cytometry (polytypic) recorded from November 2019 to December 2021 including one additional case ^d from 2022.

Diagnosis	Case	Site	Immunophenotype c	PCR analysis	Possible reason for polytypic TRBC1 pattern
AITL (Figure S2)	1	Bone marrow	CD3+, CD2+, CD4+, CD5+, CD7+, PD1+, BCL6+, CD10−	Monoclonal TCR β gene rearrangement	Few lymphoma cells
ALCL, ALK+	2	Inguinal lymph node	CD3−, CD30+, ALK+	No PCR	Few CD30+ lymphoma cells
ALCL, ALK− (Figure S3)	3	Axillary lymph node	CD3+, CD30+, CD2+, CD4+, CD5+, CD7−, ALK−	No monoclonal gene rearrangement of TCR γ or TCR β	Few CD3+/CD30+ lymphoma cells
ALCL, ALK−	4	Bone marrow	CD3− a , CD5+, CD2+, CD30+, ALK−	No PCR	Few CD30+ lymphoma cells
ALCL, ALK−	5	Axillary lymph node	CD3−, CD30+, CD2+, CD4+, CD5−, CD7+, ALK−	No PCR	CD3‐negative lymphoma cells; CD30‐staining not done on flow
ALCL, ALK+	6	Neck lymph node	CD3−, CD30+, CD2+, ALK+	No PCR	CD3‐negative lymphoma cells; CD30‐staining not done on flow
AITL (Figure 2)	7	Axillary lymph node	CD3+, CD2+, CD4+, CD5+, CD7+, PD1+, BCL6+, CXCL13+, CD10−	Monoclonal TCR γ and TCR β gene rearrangements	Unclear
T‐LGLL (Figure 3)	8	Bone marrow	CD3+, CD4+, CD8+, CD57+	Monoclonal TCR γ and TCR β gene rearrangements	Unclear
MF b	9	Blood	CD3+, CD4+, CD26−	Monoclonal TCR γ gene rearrangement	Unclear
MF b	10	Blood	CD3+, CD4+, CD26−	No PCR	Unclear
PTCL NOS d	11	M gluteus maximus	CD3+, CD2+, CD4+, CD5+, CD7−, PD1+, BCL6+, CXCL13−, CD10−	Monoclonal TCR γ and TCR β gene rearrangements	Unclear

Abbreviations: AITL, angioimmunoblastic T‐cell lymphoma; ALCL, ALK−, ALK‐negative anaplastic large cell lymphoma; ALCL, ALK+, ALK‐positive anaplastic large cell lymphoma; M, musculus; MF, mycosis fungoides; PTCL NOS, peripheral T‐cell lymphoma not otherwise specified; T‐LGLL, T‐cell large granular lymphocytic leukemia.

^a Cytoplasmatic CD3‐staining was positive by flow cytometry.

^b These two cases were from the same patient.

^c For Cases 1–7 and 11, the immunophenotype of the lymphoma cells was determined by immunohistochemistry. For Cases 8–10, including the polytypic T‐LGLL cells and MF (Sézary) cells, the immunophenotype of the neoplastic T‐cells was inferred from flow cytometry.

^d Additional case from 2022.

In a patient with BM involvement by AITL (Case 1, Table 3) the neoplastic T‐cells were not seen by MFC with only 10% T‐cells with polytypic expression of TRBC1 (Figure S2a). A prominent histiocytic inflammatory reaction and fibrosis were seen in the BM biopsy (Figure S2b). A monoclonal TCR β gene rearrangement could be demonstrated by PCR performed on DNA isolated from the BM biopsy material (Figure S2c), identical to that of a concomitant biopsy of an axillary lymph node (Figure S2d). Thus, the material analyzed by MFC was probably not representative. In another case (Case 3, Table 3), the initial FNA of an axillary lymph node showed polytypic expression of TRBC1 (Figure S3a), despite the presence of lymphoma cells on cytology (Figure S3b). Histology and immunohistochemistry revealed sheets of large CD3+/CD30+ lymphoma cells thus establishing the diagnosis of ALK‐negative ALCL a few weeks later (Figure S3c). No monoclonal TCR gene rearrangement could be demonstrated by PCR (Figure S3d) on the histologic material. Here, the possibility of rearrangements that are not covered by BIOMED primers cannot be excluded [21].

For the remaining cases, the lack of TRBC1 restriction was more difficult to explain. A case of AITL in the axillary lymph node (Case 7, Table 3) with 54% CD3+/CD4+ T‐cells displayed a polytypic pattern with the TRBC1 antibody (Figure 2a). Nevertheless, morphology and immunohistochemistry were consistent with AITL (Figure 2b) and PCR showed monoclonal gene rearrangements of TCR γ (Figure 2c) and TCR β (not shown). In another patient suffering from advanced T‐LGL leukemia for several years (Case 8, Table 3, and Figure 3a), monoclonal gene rearrangements of the TCR β gene and TCR γ could be demonstrated by PCR (not shown) despite the absence of TRBC1 restriction by MFC (Figure 3b). In these two cases, reactive polytypic T‐cell populations with immunophenotypes overlapping with the lymphoma cells could be present.

FIGURE 2.

Patient with angioimmunoblastic T‐cell lymphoma in axillary lymph node without monotypic T‐cells. (a) Polytypic expression of TRBC1 by flow cytometry. The CD3 versus CD5 plot is from the screening panel as described in Materials and Methods and shows all CD3+ or CD5+ T‐cells. Light blue cells are CD4+ T‐cells and dark blue CD8+ T‐cells. (b) Axillary lymph node biopsy stained with Hematoxylin–Eosin, for CD3, PD1, and CXCL13. (c) Monoclonal TCR γ gene rearrangement by PCR. CD, cluster of differentiation; CXCL13, C‐X‐C motif chemokine ligand 13; PCR, polymerase chain reaction; PD1, programmed cell death protein 1; TCR, T‐cell receptor; TRBC1, constant region 1 of T‐cell receptor β.

FIGURE 3.

Patient with clinically advanced T‐cell large granular lymphocytic leukemia in bone marrow without monotypic T‐cells. (a) Bone marrow biopsy stained with Hematoxylin–Eosin and for CD3. (b) Polytypic expression of TRBC1 on CD4+/CD57−, CD4+/CD57+, CD8+/CD57−, and CD8+/CD57+ subsets of CD3+ T‐cells by flow cytometry. The CD3 versus CD5 plot is from the screening panel as described in Materials and Methods and shows all CD3+ or CD5+ T‐cells. Light blue cells are CD4+ T‐cells and dark blue CD8+ T‐cells. CD, cluster of differentiation; TRBC1, constant region 1 of T‐cell receptor β.

In two consecutive samples from a patient with MF with blood involvement (1%–5% CD3+/CD4+/CD26+/CD7 heterogenous cells), no monotypic T‐cells could be found despite clear‐cut monoclonal gene rearrangement of TCR γ identical to that of the primary tumor in the skin (Cases 9 and 10, Table 3) but no monoclonal TCR β gene rearrangement. Finally, TRBC1 restriction could not be demonstrated in a patient with manifestation of PTCL NOS (Case 11, Table 3) in a gluteal muscle biopsy despite histology, immunophenotype and monoclonal gene rearrangements of both TCR β and TCR γ compatible with PTCL NOS (Table 3).

3.3. TRBC1 Monotypic Cases Not Fulfilling Criteria for T‐Cell Lymphoma

Explaining the modest PPV of TRCB1 restriction for T‐cell neoplasms, we found 48 (56 with surface CD3‐negative cases included) cases (32.9% of all TRBC1 restricted cases) with small monotypic T‐cell populations without definite evidence of T‐cell lymphoma or fulfilling diagnostic criteria for T‐LGL leukemia (Table 1). These cases were designated (1) T‐cell clones of uncertain significance related to T‐LGL proliferations (T‐CUS‐LGL), if they displayed a T‐LGL phenotype without fulfilling criteria for T‐LGL leukemia [16, 25], (2) T‐cell clones of uncertain significance related to classic Hodgkin lymphoma (T‐CUS‐cHL), if they were found in a lymph node with cHL, or (3) T‐cell clones of uncertain significance, not otherwise specified (T‐CUS NOS), if they did not fulfill the criteria for T‐CUS‐LGL or T‐CUS‐cHL (Table S1).

For 79 of the monotypic TRBC1 cases PCR was performed, confirming monoclonality in 61 cases (Table 2). Of the 18 cases with TRBC1 restriction and polyclonal PCR results, 9 showed low levels of monotypic T‐cells (< 3%) (Table S1) offering a likely explanation for the lack of monoclonal TCR gene rearrangement. In five of the remaining cases (displaying 5%–12% monotypic T‐cells), PCR results were in retrospect found to be ambiguous with oligoclonal rather than clear‐cut polyclonal results.

3.4. Immunophenotype of TRBC1 Restricted T‐Cell Populations

Among the 146 monotypic T‐cell proliferations (Table 1), 79 cases expressed and 67 lacked TRBC1 (Table S1). Most of the monotypic CD3+ T‐cell populations were CD4+ (Table S1). In general, the fraction of monotypic T‐cells was higher in T‐cell neoplasms than in T‐CUS, with the exception of T‐follicular helper cell lymphomas, including AITL and nodal peripheral T‐cell lymphoma with T‐follicular helper phenotype (Table S1). T‐CUS LGLs closely resembled the immunophenotype of T‐LGLL such as loss of CD7 or dim expression of CD5 and were most often CD8+/CD4− (56.7%) although a substantial subfraction (26.7%) was CD4+/CD8− (Table S3). However, there was no significant difference between T‐LGLL and T‐CUS LGL with respect to the CD8+/CD4− or CD4+/CD8− immunophenotype (Table S3).

Apart from the T‐CUS LGL cases, we identified 22 T‐CUS NOS and 4 T‐CUS‐cHL cases, of which 25 (96%) exhibited an aberrant immunophenotype (Table S4). Fifteen T‐CUS NOS cases (58%), originating from 10 patients, were associated with unrelated malignancies, and two patients with T‐CUS NOS had inflammatory conditions (liver cirrhosis and psoriasis) (Table S4). The TRBC1‐restricted T‐cell neoplasms lacking TRBC1 more often displayed a cytotoxic phenotype (CD8+/CD4−) than those expressing TRBC1 (p = 0.02; Table S3). By contrast, there was no statistically significant difference between T‐CUS TRBC1+ and TRBC1− regarding the CD8+/CD4− phenotype (Table S3).

3.5. Follow‐Up of T‐CUS Cases

To further elucidate the clinical relevance of T‐CUS, we analyzed follow‐up data on the evolution of clone size over time, as determined by the percentage of TRBC‐restricted lymphocytes. Flow cytometric data were available for 24 of the 56 cases (43%), with one to three follow‐up time points ranging from 1 to 77 months (Table S5). Of these, 19 cases remained stable, while four T‐CUS‐LGL cases and one T‐CUS‐cHL case showed an increase in clone size, although without progression to T‐cell lymphoma (Table S5).

4. Discussion

Determination of T‐cell monoclonality by TRBC1 restriction (monotypic presence or absence) on flow cytometry has rapidly emerged as a powerful tool to discriminate neoplastic T‐cell clones from reactive or benign T‐cell populations. We confirm the high sensitivity of the method although we found some T‐cell lymphomas without clear restriction of TRBC1, including CD3‐expressing ALCL and two cases of AITL. In addition, we found suspect surface CD3+ T‐cell clonal populations without evidence of T‐cell neoplasm related to small stable T‐LGL‐clones (T‐CUS‐LGL), clonal expansions of reactive T‐cells in classic Hodgkin lymphoma (T‐CUS‐cHL) or other clones (T‐CUS NOS) challenging the high specificity claimed by some reports [6]. We did not assess the cytoplasmic expression of TRBC1, which would have been valuable in further determining the accuracy of the method, as T‐cell lymphomas are sometimes surface CD3‐negative/CD5‐positive.

Previous studies reported almost perfect concordance between monotypic T‐cell populations and T‐cell neoplasm regardless of affected tissue, reaching 97%–100% sensitivity in several studies [6, 11, 12, 13]. The reported specificity is also high; 84%–100% depending on the tissue examined; specificity is considerably lower in blood or BM samples, largely due to the frequent finding of small, most often CD8+ T‐cell clones of unknown significance (T‐CUS) that do not equal malignancy [12, 13]. Yet, in lymph nodes or other tissues, including body fluids, the specificity appears to be very high, reaching 100% in one report [6].

TRBC1‐restriction analysis displayed an excellent ability to pick up neoplastic T‐cell clones in our hands, reaching a sensitivity of 90.7% albeit with a lower PPV of 67.1% (Table 1). The overall PPV must be interpreted with some caution, as TRBC1 restriction analysis was performed on some BM and PB samples without suspicion of T‐cell lymphoma, including patients with pancytopenia (see Materials and Methods). Nevertheless, PPV also proved to be quite modest (70.6%) when TRBC1 restriction was analyzed on FNA and body fluids alone (Table S2). The more recent addition of TRBC2 antibodies has improved sensitivity and specificity, particularly in previously equivocal cases with dim TRBC1 expression. Based on TRBC1/TRBC2 ratios, all these cases could be resolved into unequivocal restricted/malignant or nonrestricted/reactive phenotypes [26]. If flow cytometry is unavailable, immunohistochemistry for TRBC1 may serve as a valid and practical alternative to identify monoclonal T‐cell populations in tissue sections [27].

Although the results from our everyday hematopathology practice clearly show the benefit of including the TRBC1 antibody in a routine T‐cell panel thus reducing the likelihood of a T‐cell neoplasm to ~1% with polytypic expression of TRBC1, some polytypic cases proved to be T‐cell lymphomas. A comprehensive TRBC1 restriction analysis by MFC requires careful assessment of all small subsets of aberrant or rare CD3+ T‐cell populations, but even with careful retrospective analysis of the polytypic T‐cell lymphoma cases, we could not identify monotypic T‐cells in any of them except one case of AITL (Case 1, Table 3, Figure S2).

The inflammatory component, which certainly also includes T‐cells, may be considerable in some T‐cell lymphomas, especially AITL where the neoplastic T cells often constitute a minor part of the cellular infiltrate and may be difficult to identify [28, 29]. Nevertheless, AITL cells are often surface CD3−/CD5+ which makes them easier to identify even with minute amounts of lymphoma cells [30]. Another possible explanation for the lack of TRBC1 restriction in some T‐cell lymphomas could be the presence of two or more subsets of neoplastic T‐cells in AITL, but the detection of single monoclonal TCR gene rearrangements by PCR, such as in Case 7 (Table 3 and Figure 2), argues against this interpretation. It is more tempting to speculate that most cells in such cases are inflammatory (polyclonal) rather than neoplastic (monoclonal).

We found five ALCLs without monotypic expression of TRBC1, making this entity the most frequently missed by MFC. Paucity of intact lymphoma cells in the investigated sample may have contributed to the inability to detect some of these cases. In such cases, the polytypic TRBC1 expression of the remaining background non‐neoplastic T‐cells can be misleading. Of course, clonal analysis of surface TRBC1 expression is not helpful in surface CD3‐negative T‐cell lymphomas including some cases of ALCL, AITL or T‐lymphoblastic lymphoma because of the inherent lack of TCR β chains in CD3‐negative cells [31]. These lymphomas are instead detected by MFC precisely because of the lack of surface CD3 in combination with the expression of other T‐cell markers. This implies careful analysis of surface CD3−/CD5+ (CD2+/CD7+) populations by flow cytometry.

The 10 T‐cell neoplasms without detectable TRBC‐restricted subsets by flow cytometry of the present study illustrate the strength of two independent methods to determine clonality status (except for one case of ALK‐negative ALCL), defying detection of monoclonal expansion by both MFC and PCR (Case 3, Table 3, and Figure S3). If MFC is inconclusive or shows ambiguous results, a morphological suspicion of T‐cell lymphoma can usually be confirmed with PCR.

Similarly to a previous study [12], we found 56 cases with monotypic T‐cells without evidence of T‐cell neoplasm. In contrast to that study, we did not use TRBC restriction analysis as a screening tool. Instead, TRBC1 expression was assessed when there was clinical or morphological suspicion of T‐cell lymphoma, or when we aimed to exclude such a diagnosis. Notably, about one third of the cases showing TRBC restriction ultimately did not represent T‐cell lymphomas. In the study by Shi et al., most clones were T‐LGL‐like, often found in healthy donors without evidence of T‐cell neoplasm and designated T‐CUS [12]. In our material we noted several TRBC1 restricted T‐LGL‐like clones (T‐CUS‐LGL), albeit at a lower frequency. In the present study, we applied a cut‐off of 0.5 × 10⁹/L monotypic LGL cells in at least two consecutive samples > 6 months apart to define LGL‐leukemia, which is a commonly accepted threshold [25, 32, 33]. Clearly, TRBC1 restriction analysis may eliminate the need for molecular clonality testing in the context of LGL leukemia and circulating disease in patients with cutaneous T‐cell lymphoma [34]. Interestingly, we found a much higher frequency (26.7%) of CD4+/CD8‐T‐CUS‐LGL populations than the modest 2% reported in a previous study [12]. A possible explanation could be different inclusion criteria; in the cited study, patients were selected based on the absence of clinical or laboratory evidence of a current, prior, or subsequently diagnosed T‐cell malignancy. We could not confirm the previously reported slightly dimmer expression of CD2 and CD7 in T‐LGLL compared to T‐CUS‐LGL [12].

We also discovered small T‐cell clones without LGL cell phenotype (T‐CUS NOS and T‐CUS‐cHL) in specimens from various tissues, almost all with immunophenotypical aberrancies, perhaps reflecting an immunologic response to the tumor or the specific inflammatory disorder. T‐CUS, beyond cases with a T‐LGL immunophenotype, has not been studied extensively before. Pu et al. described 15 cases of which at least one third showed immunophenotypic aberrancies apparently unrelated to T‐LGL [35]. These T‐CUS cases were most often found in the context of another unrelated primary tumor. In another recent study, cCD3‐negative T‐CUS was described resembling AITL [36]. Together, these data suggest that phenotypical aberrancies in T‐CUS are common. Interestingly, in lymph nodes from four cases of cHL (with B‐cell phenotype of the Hodgkin Reed Sternberg cells), a concomitant monotypic T‐cell population was found of which three were immunophenotypically aberrant (Table S4). PCR analysis was available from two of these cases, one without and one with monoclonal TCR gene rearrangement, probably reflecting the monotypic expansion of T‐cells by flow cytometry. Monoclonal TCR γ gene rearrangements may occur in rare cases of cHL, sometimes reflecting the very unusual T‐cell phenotype of Reed Sternberg cells [37], which is unlikely to be the case here, as none of the Hodgkin Reed Sternberg cells of the cHL's displayed specific T‐cell markers. Clonal T‐cell expansion in cHL has also been described, although almost exclusively restricted to CD8+ T cells in contrast to our cases, which were all CD4+ [38]. Another earlier study failed to demonstrate clonal expansion of T‐cells in cHL [39].

5. Conclusion

We corroborate the high accuracy of TRBC1 restriction analysis to detect monotypic T‐cells with high concordance with PCR results for monoclonal T‐cell receptor gene rearrangements. It is a highly sensitive assay for detection of T‐cell neoplasms although ALCL and other T‐cell lymphomas with a significant inflammatory component such as AITL may escape detection. It is also highly specific but needs to be interpreted with caution since monoclonal expansion of T‐cells can be found in a variety of clinical contexts and tissues without bona fide T‐cell neoplasm.

Author Contributions

Conceptualization: Mats Ehinger, Anna Porwit, Olof Axler, and Nilofar Rajabian. Methodology: Olof Axler. Validation: Olof Axler. Formal Analysis: Nilofar Rajabian and Olof Axler. Investigation: Nilofar Rajabian and Olof Axler. Data curation: Nilofar Rajabian and Erik Wistén. Writing – original draft preparation: Nilofar Rajabian, Mats Ehinger. Writing – review and editing: Nilofar Rajabian, Anna Porwit, Erik Wistén, Olof Axler, and Mats Ehinger. Visualization: Nilofar Rajabian and Mats Ehinger. Supervision: Mats Ehinger. Funding Acquisition: Mats Ehinger.

Funding

This work was supported by Region Skåne UFo grants, Governmental Funding of Clinical Research (ALF) and Lund University medical faculty funds.

Ethics Statement

Ethical approval for the study was granted by the Swedish Ethical Review Board (No. 22‐02948‐01).

Consent

The authors have nothing to report.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Figure S1: Gating strategy for TRBC restriction (a case of PTCL NOS in peripheral blood). The starting point in the analysis is to gate all CD3+ cells in the screening tube, excluding negative cells and debris (not shown). In the first step (a), the CD4/CD8 ratio is determined followed by (b) analysis of CD5 expression (negative on the lymphoma cells in this case). Next, the tube containing the extended T‐cell panel including TRBC1 is analyzed. After exclusion of TCR γδ + cells (not shown), expression of TRBC1 is examined on CD4+ T‐cells (c), CD7+ subsets (d), CD30+ subsets (e), CD2+ subsets (not shown) and CD8+ T‐cells (f). Hence, the lymphoma cells in this case were CD3+/CD4+/CD8‐/CD2+/CD5−/CD7−/CD30−/TCR γδ− (a–e) with TRBC1‐restriction (c–e; > 85% TRBC1+/CD4+ T‐cells). In the background, normal populations of CD3+/CD4+/CD5+/CD7+ (a–e) and CD3+/CD8+ (a and f) T‐cells can be seen with polytypic expression of TRBC1 (d and f). CD, cluster of differentiation; TCR, T‐cell receptor; TRBC1, constant region 1 of T‐cell receptor β.

Figure S2: Patient with angioimmunoblastic T‐cell lymphoma in bone marrow without monotypic T‐cells. (a) Polytypic expression of TRBC1 by flow cytometry. The CD3 versus CD5 plot is from the screening panel as described in Materials and Methods and shows all CD3+ or CD5+ T‐cells. Light blue cells are CD4+ T‐cells and dark blue CD8+ T‐cells. (b) Bone marrow biopsy stained with Hematoxylin–Eosin, for CD3 and for CD163. (c) Monoclonal TCR β gene rearrangement by PCR in bone marrow. (d) Identical monoclonal TCR β gene rearrangement by PCR in axillar lymph node with lymphoma. CD, cluster of differentiation; PCR, polymerase chain reaction; TCR, T‐cell receptor; TRBC1, constant region 1 of T‐cell receptor β.

Figure S3: Patient with ALK‐negative anaplastic large cell lymphoma in axillary lymph node without monotypic T‐cells. (a) Absence of CD3+/CD4+/CD30+ lymphoma cells with polytypic expression of TRBC1 on CD3+/CD4+/CD5+ T‐cells by flow cytometry on axillary lymph node fine needle aspirate. The CD3 versus CD5 plot is from the screening panel as described in Materials and Methods and shows all CD3+ or CD5+ T‐cells. Light blue cells are CD4+ T‐cells and dark blue CD8+ T‐cells. (d) Cytology (Giemsa stain) on fine needle aspirate from axillary lymph node revealing a few very large lymphoma cells. (e) Histology (Hematoxylin–Eosin) from axillary lymph node revealing sheets of large CD30+ lymphoma cells. (f) Absence of monoclonal TCR γ rearrangements by PCR in axillary lymph node. CD, cluster of differentiation; PCR, polymerase chain reaction; TCR, T‐cell receptor; TRBC1, constant region 1 of T‐cell receptor β.

Table S1: TRBC1 expression, sites of involvement, CD4/CD8 ratio and range of monotypic T‐cells of all consecutive TRBC1‐interrogated T‐cell lymphoproliferative disorders (n = 183) retrieved from November 2019 to December 2021 comprising 127 bona fide T‐cell neoplasms, 48 monotypic CD3+ T‐cell expansions of uncertain significance and 8 CD3− T‐cell expansions. AITL, angioimmunoblastic T‐cell lymphoma; ALCL, ALK−, ALK‐negative anaplastic large cell lymphoma; ALCL, ALK+, ALK‐positive anaplastic large cell lymphoma; ATL, adult T‐cell leukemia/lymphoma; MF/SS, mycosis fungoides/Sézary syndrome; Nodal PTCL‐TFH, Nodal peripheral T‐cell lymphoma with T‐follicular helper phenotype; N/A, not applicable; PTCL NOS, peripheral T‐cell lymphoma, not otherwise specified; TCL NOS, T‐cell lymphoma, not otherwise specified; T‐CUS‐cHL, T‐cell clones of uncertain significance in classic Hodgkin lymphoma; T‐CUS‐LGL, T‐cell leukemia clones of undetermined significance with large granular lymphocytes; T‐CUS NOS, T‐cell clones of undetermined significance, not otherwise specified; T‐LBL, T‐lymphoblastic leukemia/lymphoma; T‐LGLL, T‐cell large granular lymphocytic leukemia; T‐PLL, T‐cell prolymphocytic leukemia.

Table S2: Diagnostic performance of TRBC1 restriction analysis by flow cytometry in 188 consecutive cytology samples (184 fine needle aspirations, four pleural effusions, and four cerebrospinal fluid samples) extracted from Table 1 with or without T‐cell lymphoma (excluding gamma/delta T‐cell neoplasms) excluding or including surface CD3‐negative (sCD3−) cases. NPV, negative predictive value; PPV, positive predictive value.

Table S3: Expression of CD4 and CD8 in all CD3+ TRBC1‐restricted T‐cell neoplasms and all TRBC1‐restricted CD3+ T‐cell clones of undetermined significance (T‐CUS) analyzed by multiparameter flow cytometry. In addition, all T‐LGLLs (T‐cell large granular lymphocytic leukemia) and all T‐CUS LGLs (T‐cell leukemia clones of undetermined significance with large granular lymphocytes) were analyzed separately (bottom two rows).

Table S4: Immunophenotypes and sites of T‐CUS NOS and T‐CUS‐cHL cases. BM, bone marrow; Dim, weak expression; DLBCL, diffuse large B‐cell lymphoma; hetero, heterogenous expression; MDS‐EB1, myelodysplastic syndrome excess blasts type 1; PB, peripheral blood; T‐PLL, T‐cell prolymphocytic leukemia; ++, strong expression; †, ‡, §, ¶, #, and * denote cases belonging to the same patient. #, this patient had two different T‐CUS‐cHL clones present at both indicated sites.

Table S5: Follow‐up data on T‐CUS patients (n = 24). Clone size as determined by % TRBC‐restricted lymphocytes of all lymphocytes at indicated time intervals (months) after initial diagnosis. The site is peripheral blood unless otherwise indicated. BM, bone marrow; LN, lymph node; †, cases with progression of T‐CUS over time.

Rajabian N., Axler O., Wistén E., Porwit A., and Ehinger M., “Restricted Expression of the Constant Region 1 of T‐Cell Receptor β by Flow Cytometry Facilitates Detection of T‐Cell Neoplasms With High Specificity but Moderate Predictive Value,” European Journal of Haematology 116, no. 3 (2026): 245–255, 10.1111/ejh.70067.

Data Availability Statement

Data is available at the Department of Clinical Pathology, Lund, Regional Laboratories Region Skåne, Lund, Sweden.