How I Diagnose Angioimmunoblastic T-Cell Lymphoma

Yi Xie, MD, PhD; Elaine S. Jaffe, MD


Am J Clin Pathol. 2021;156(1):1-14. 

In This Article

Questions for the Experts

1. How do we distinguish AITL with a type I pattern from reactive paracortical hyperplasia?

A significant proportion of AITL cases (up to 17%) present histologically with the type I pattern (also called "early" pattern).[29,30] The distinction between type I AITL and reactive paracortical (T-zone) hyperplasia can be challenging due to their overlapping morphologic features.[29] As mentioned earlier, AITL pattern I usually comprises hyperplastic follicles with ill-defined borders and attenuated mantle zones. While the presence of atypical clear cells at the outer rim of the germinal centers favors AITL, in many cases, the atypical cells may be inconspicuous, and morphologic findings alone are insufficient to signify a correct diagnosis.

Immunohistochemical staining for the TFH cell markers, particularly PD1, is of great value in distinguishing AITL type I with reactive paracortical hyperplasia. In paracortical hyperplasia, strong PD1-positive cells are confined to the germinal centers (primarily to the light zone at the periphery of germinal centers), although there may be scattered or sometimes numerous extrafollicular T cells with variable and weak PD1 positivity.[40] In contrast, PD1 demonstrates a strong perifollicular staining pattern in AITL, corresponding to the "perifollicular" distribution of neoplastic TFH cells. Markers such as CD10, if strongly expressed in perifollicular T cells, are a helpful clue. Expansion of CD21+ meshworks with a perivascular distribution, seen in pattern II or III, is usually not evident in pattern I. The presence of interfollicular EBV-positive B blasts is also helpful but may not be a distinguishing feature, as it can also be seen in acute and chronic EBV infection as well as other lymphoproliferative disorders.

Clinical features have particular importance in the diagnosis of AITL. Interestingly, despite early involvement (pattern I) histologically, many patients reportedly had clinically advanced disease.[15] Caution should be exercised in cases that lack characteristic clinical symptoms. If an early AITL is suspected, clonality studies should be obtained. The findings of clonal T-cell gene rearrangement, with or without concurrent B-cell gene rearrangement, are helpful in confirming a diagnosis of AITL. As always, the final diagnosis should be based on both the clinical and pathologic features as well as the molecular findings. In equivocal cases, close follow-up and repeat biopsy are recommended to establish a definitive diagnosis.

2. Where should we draw the line between B-cell proliferation in AITL and so-called composite lymphoma?

As mentioned previously, the presence of large B cells, often with immunoblastic features, usually but not always infected by EBV, is a characteristic feature of AITL.[46] These cells are typically scattered in the interfollicular zone, sometimes showing an HRS-like appearance.[27] However, in some cases, there may be prominent expansion of EBV-positive B cells, forming large clusters to confluent foci of large transformed B cells, focally obscuring the coexisting T-cell component.[44] B-cell proliferation may progress, and so-called secondary large B-cell lymphomas have been reported in a few cases, occurring simultaneously or before the initial diagnosis of AITL or, more commonly, at relapse.[44,45] Moreover, there are rare reports of AITL and DLBCL in the same anatomic site, so-called composite AITL and large B-cell lymphoma.[78,79] Our own clinical experience suggests that that prognosis is dominated by AITL and that the EBV-positive B-cell expansions are less significant clinically. Given the B-cell expansion in AITL, rituximab had been incorporated into clinical treatment protocols, but long-term clinical benefit was not shown for this approach.[80] Therefore, a diagnosis of "composite" lymphoma is generally not made in these cases. Recent data have found shared mutations within the B cells and T cells in some cases of AITL, raising the possibility that AITL arises in a pluripotential stem cell. These observations are provocative but require further investigation to decipher the clonal relationship of the B cells and T cells within AITL.

3. How many TFH markers are needed for the diagnosis of AITL, and which ones need to be done routinely?

Since being recognized in the 1970s, several immunohistochemical and flow cytometric markers have been reported that could aid in the identification of neoplastic TFH cells in AITL. These include surface markers PD1, CD10, CD200, ICOS, and cytoplasmic SAP; transcription factor BCL6 and c-MAF; the chemokine CXCL13 and its receptor CXCR5; and so on.[9,29,35–39] Among TFH markers, PD1 and ICOS are more sensitive for identifying the neoplastic TFH cells, whereas CXCL13 and CD10 are more specific.[42] However, none of the TFH markers in isolation is 100% sensitive or specific for the TFH phenotype. For this reason, the 2016 WHO requires demonstrating expression of at least two but ideally three TFH markers to establish a diagnosis of TFH lymphoma.[14] We routinely perform CD10, BCL6, ICOS, and PD1 stains on all cases and additional TFH markers such as CXCL13 when necessary. Basha et al[81] recently evaluated the utility of a five-marker TFH panel (CD10, BCL6, PD1, CXCL13, and ICOS) for the diagnosis of AILT and PTCL-TFH. They showed that while a four-marker panel (CD10, BCL6, PD1, CXCL13) was adequate to diagnose most AITLs, a five-marker panel with the addition of ICOS significantly increased the ability to identify PTCL-TFH. A significant issue is intensity, since many of these markers (PD1, ICOS) can be dimly positive, a feature with less specificity. With CD10, the number of positive cells may be only few, but if they are cytologically atypical, even few positive cells are relevant. It is important to bear in mind that AITL is based on a constellation of clinical and pathologic features. In a classic case, with expansion of FDCs around the HEVs, arborizing vascular proliferation, and increased EBV-positive cells, fewer TFH markers may be required to reach a confident diagnosis of AITL.

4. How do we differentiate AITL and other PTCL of TFH derivation with HRS-like cells from classic Hodgkin lymphoma?

As mentioned earlier, HRS-like cells are common findings in lymphomas of TFH derivation, such as AITL and FTCL, and can assume both the morphology and immunophenotype of classic Reed-Sternberg cells of Hodgkin lymphoma.[46] As atypia in the background T-cell population may be minimal, a potential pitfall in the differential diagnosis of the lymphomas of TFH derivation is misdiagnosing CHL. This is especially so in FTCL, in which HRS-like cells are present in a background of small lymphocytes with a nodular growth pattern, closely mimicking lymphocyte-rich CHL.[82]

Like CHL, the HRS-like cells in PTCL-TFH are consistently positive for CD30 and show weak PAX5, often with coexpression of CD15, negative or variably positive for CD20. In PTCL-TFH, the HRS-like cells are commonly EBV infected, and in situ hybridization for EBER most often highlights a wider range of positive cells than seen in EBV-positive CHL; however, this is not always the case. The key to the correct diagnosis lies in the recognition of a background atypical T-cell component with appropriate immunophenotypic and molecular studies. In FTCL, immunostaining characteristically highlights aggregates of small- to medium-sized, mildly atypical T cells with a TFH immunophenotype, surrounding the HRS-like cell. In contrast, TFH markers usually identify a single layer of reactive T cells forming rosettes around HRS cells in lymphocyte-rich CHL.[46,82] Accordingly, immunophenotyping by flow cytometry frequently detects a distinct CD3−/dim CD4+ T-cell population in AITL and FTCL but not in CHL, which may also provide a clue in the differential diagnosis.[34] In some cases of PTCL-TFH, the neoplastic T cells may also express variable CD30, whereas CD30 staining is typically restricted to the HRS cells in CHL. A common finding is clusters of atypical CD30+ or MUM1+ T cells surrounding the HRS-like cells. The findings of open peripheral sinuses, prominent arborizing HEVs, and extrafollicular FDC expansions are helpful features to suggest a diagnosis of AITL rather than CHL. While the correct diagnosis can generally be made based on morphologic and immunophenotypic findings, in difficult cases, molecular studies may be necessary to confirm the diagnosis.