Parallel evolution of two distinct lymphoid proliferations in clonal haematopoiesis

Abstract

Aims: Angioimmunoblastic T-cell lymphoma (AITL) is genetically characterized by TET2 and DNMT3A mutations occurring in haematopoietic progenitor cells, and late events (e.g. the RHOA-G17V mutation) associated with malignant transformation. As TET2/DNMT3A-mutated progenitor cells can differentiate into multilineage progenies and give rise to both AITL and myeloid neoplasms, they may also have the potential to lead to other metachronous/synchronous neoplasms. We report two cases showing parallel evolution of two distinct potentially neoplastic lymphoid proliferations from a common mutated haematopoietic progenitor cell population.

Methods and results: Both cases presented with generalized lymphadenopathy. In case 1 (a 67-year-old female), an initial lymph node (LN) biopsy was dismissed as reactive, but a repeat biopsy showed a nodal marginal zone lymphoma (NMZL)-like proliferation with an increase in the number of T-follicular helper (TFH) cells. Immunohistochemistry, and clonality and mutational analyses by targeted sequencing of both whole tissue sections and microdissected NMZL-like lesions, demonstrated a clonal B-cell proliferation that harboured the BRAF-G469R mutation and shared TET2 and DNMT3A mutations with an underlying RHOA-G17V-mutant TFH proliferation. Review of the original LN biopsy showed histological and immunophenotypic features of AITL. In case 2 (a 66-year-old male), cytotoxic T-cell lymphoma with an increase in the number of Epstein-Barr virus-positive large B cells was diagnosed on initial biopsy. On review together with the relapsed biopsy, we identified an additional occult neoplastic TFH proliferation/smouldering AITL. Both T-cell proliferations shared TET2 and DNMT3A mutations while RHOA-G17V was confined to the smouldering AITL.

Conclusions: In addition to demonstrating diagnostic challenges, these cases expand the potential of clonal haematopoiesis in the development of different lineage neoplastic proliferations.

Keywords: TET2 and DNMT3A mutation; angioimmunoblastic T-cell lymphoma; clonal haematopoiesis; secondary lymphoid neoplasm.

Figure 1

Histopathological and immunohistochemical investigations of lymph node (LN) biopsies in case 1. A, A cervical LN showing a florid B‐cell and plasmacytic infiltrate mimicking nodal marginal zone lymphoma (NMZL)/secondary NMZL, obscuring underlying angioimmunoblastic T‐cell lymphoma (AITL). Upper panels: on an haematoxylin and eosin (H&E)‐stained section (left), there is effacement of the nodal architecture by an expanded perifollicular infiltrate (right) composed of small and medium‐sized lymphocytes and mature plasma cells. Middle panels: CD79a staining highlights the B cells and plasma cells (left), whereas CD138 marks the dense clusters of mature plasma cells (right). Lower panels: there is an increase in the number of programmed cell death protein 1 (PD1)‐positive T cells within the germinal centres (strong staining) and also in the perifollicular areas (left), but CD21 staining does not show any extrafollicular follicular dendritic cell (FDC) expansion (inset); a clonal B‐cell population in a cervical LN biopsy was detected by polymerase chain reaction with the BIOMED‐2 assay (IGH FR3‐JH) (right). B, A groin LN showing typical features of AITL. Upper panels: on an H&E‐stained section (left) there is striking high endothelial venule (HEV) hyperplasia, amidst which is an infiltrate of medium‐sized atypical lymphoid cells in a background of histiocytes, plasma cells, and eosinophils. On a CD21‐stained section (right), there is prominent FDC expansion, seen to encircle HEVs. Lower panels: the atypical lymphoid cells are strongly positive for PD1 (left), and a proportion are positive for CD10 (left side of the right panel) and CXCL13 (right side of the right panel).

Figure 2

Genetic findings in case 1. A, A summary of the mutations detected in various lymph node (LN) biopsies and blood samples by the use of targeted next‐generation sequencing and quantitative polymerase chain reaction (PCR) (qPCR). The allelic frequencies of the variants included are based on Fluidigm PCR and MiSeq sequencing unless otherwise indicated. *Based on Royal Marsden Hospital (RMH) panel sequencing. The sequencing coverage for all the variant loci by the Fluidigm PCR and MiSeq approach is adequate (mean of 976 reads and range of 118–1617 reads passing sequence quality checks), as PCR primers are designed to amplify short sequence fragments. cfDNA, cell‐free DNA; PBMC, peripheral blood mononuclear cell. B, Examples of RHOA‐G17V mutation analysis by qPCR. The qPCR analysis of whole tissue sections of the groin LN was based on crude DNA preparations with varying amounts of DNA for different replicates, thus giving rise to different profiles, albeit they were all positive. The qPCR analysis of other samples shown in this panel was based on purified DNA. C, Confirmation of TET2, DNMT3A and BRAF mutations in ‘nodal marginal zone lymphoma (NMZL)‐like’ B cells by microdissection and Sanger sequencing. The prominent mutant sequence traces indicate that these mutations were present in the clonal NMZL‐like B cells. The areas enriched by the B‐cell/plasmacytic NMZL‐like infiltrate but lacking prominent programmed cell death protein 1‐positive cells were first identified (marked by red line), and used to guide microdissection on consecutive tissue sections. D, Parallel evolution of a neoplastic T‐follicular helper cell proliferation/underlying AITL and an NMZL‐like proliferation from a common haematopoietic progenitor cell population that harbour TET2 and DNMT3A mutations.

Figure 3

Histopathological and genetic findings in case 2. A, Upper panel: on haematoxylin and eosin staining (left), there is a dense diffuse monotonous infiltrate of small to medium‐sized lymphoid cells. CD5 (brown)/CD8 (red) double staining showed that most of the T cells expressed CD8, but had down‐regulation/loss of CD5 expression. Lower panels: CD10 (brown)/PAX5 (red) double staining (first panel) highlights a minor population of CD10‐positive, PAX5‐negative T cells that are positive for programmed cell death protein 1 (second panel). On CD21 staining (third panel) there is mild follicular dendritic cell hyperplasia, whereas, on Epstein–Barr virus (EBV)‐encoded small RNA in‐situ hybridisation (fourth panel) there is an increase in the number of EBV‐positive cells. B, Analysis of the CD8+ T cells of the groin lymph node (LN) by microdissection and BIOMED‐2 clonality assays. The CD8+ T‐cell population (microdissected from the tissue fragment showing extensive CD8 staining but not CD4 staining, marked by a dotted black line) is clearly clonal, showing a different clonal profile from those of whole tissue sections of both groin and mediastinal LNs, but a common clonal T‐cell receptor gene rearrangement by the TRB‐B reaction, thus being present prominently in both LN specimens. C, Summary of mutations detected in the two LN specimens and microdissected CD8+ T cells. CD8+ T cells harbour the TET2 and DNMT3A mutations, but not the RHOA‐G17V mutation. The allelic frequencies of the variants included are based on Fluidigm PCR and MiSeq sequencing, and the sequencing coverage for all of the variant loci is adequate (mean of 657 reads and range of 196–1672 reads passing sequence quality checks). NA, not available.