Introduction

Chimeric antigen receptor (CAR) T-cell has emerged as a promising therapy for refractory B-cell lymphoma [1]. It improves long-term outcomes in high-grade B-cell lymphoma with MYC and BCL2 and/or BCL6 rearrangements (double-hit and triple-hit lymphoma) [2]. The treatment involves using patient-derived T cells directed against CD19 on neoplastic B cells [3]. However, long-term follow-up highlights that 60% of patients relapse [4]. Transdifferentiation events have been proposed as a means of late-onset resistance to CAR T-cell therapy [5].

Lymphoma transdifferentiation occurs when neoplastic hematopoietic cells lose the immunophenotype of one lineage and gain the immunophenotype of another, while retaining the same genetic abnormalities [6]. This phenomenon is relatively uncommon, and the mechanism is poorly understood. Histiocytic sarcoma/malignant histiocytosis (HS/MH) is a rare and unusual hematopoietic neoplasm arising from macrophages resident in tissues, termed non-Langerhans histiocytic cells. The US Surveillance, Epidemiology, and End Results (SEER) database estimates an overall incidence of 0.17 per 1,000,000 individuals [7]. HS/MH may originate sporadically as a primary malignancy (de novo). Alternatively, secondary HS/MH can occur after or occasionally simultaneously with another hematologic neoplasm, typically with evidence of a common genetic abnormality termed “clonal relationship,” such as t(14;18) seen in follicular lymphoma [8, 9]. There are several reports of secondary HS/MH occurring as a result of transdifferentiation events from low-grade B-cell lymphomas [10].

Transdifferentiation after CAR T-cell therapy is exceptionally uncommon, with rare cases reported in the literature [11, 12]. We describe a case of high-grade B-cell lymphoma with MYC and BCL2 rearrangements that underwent transdifferentiation into HS/MH following CAR T-cell therapy.

Clinical history

A 56-year-old man with a history of follicular lymphoma transformed into high-grade B-cell lymphoma with MYC and BCL2 rearrangements involving the left inguinal lymph node presented with relapsed disease. He had previously completed four cycles of DA-EPOCH-R, followed by autologous stem cell transplantation. He achieved complete clinical remission for 1 year until investigation of new right knee pain revealed an osteolytic lesion with a contiguous soft tissue mass. Biopsy of the right distal femur lesion and soft tissue revealed necrosis and lymphoid infiltrate consistent with lymphoma recurrence (Fig. 1a, b). After a short course of palliative radiotherapy, the patient underwent anti-CD19 CAR T-cell infusion, and he had tumor regression in the first 2 weeks following treatment.

Fig. 1
figure 1

Right thigh mass biopsies, pre- and post-CAR T-cell therapy. a, b Pre-CAR T-cell therapy biopsy of thigh mass revealed bone and soft tissue involved by an atypical lymphoid infiltrate (H&E, 40 ×) (a) composed of CD20-positive B cells (10 ×) (b), consistent with the previously diagnosed high-grade B-cell lymphoma with MYC and BCL2 rearrangements. c–l Post-CAR T-cell therapy biopsy of thigh mass revealed two morphologically distinct populations. B Cells are intermediate to large, with irregular nuclear contours and moderate amount of cytoplasm (H&E, 60 × oil immersion) (c). Histiocytes are large, with irregular nuclear contours and abundant eosinophilic cytoplasm (H&E, 60 × oil immersion) (d). B Cells are positive for CD19 (20 ×) (e) and CD20 (20 ×) (g), and negative for CD163 (20 ×) (i) and CD68 (KP1) (20 ×) (k). Histiocytes are negative for CD19 (20 ×) (f) and CD20 (20 ×) (h), and are positive for CD163 (20 ×) (j) and CD68 (KP1) (20 ×) (l)

Full size image

Approximately 1 month later, he presented with progressive generalized weakness, altered mental status, and rapid growth of the right thigh soft tissue mass. A follow-up PET scan showed significant FDG uptake at that site (Fig. 2) and multiple other areas of increased metabolic activity. We performed a repeat core needle biopsy of the thigh mass (Fig. 1c–l).

Fig. 2
figure 2

PET scans, pre- and post-CAR T-cell therapy. a Prior to CAR T-cell therapy, there was non-specific intermuscular uptake noted within the anterior compartment of the right thigh. b PET scan obtained 23 days after CAR T-cell infusion revealed markedly increased metabolic activity distributed throughout the right anterior thigh and right hemi-pelvis, with SUVmax of 18.3 (rated at D5b on the Deauville Five-Point Scale)

Full size image

The patient was started on polatuzumab salvage therapy for relapsed disease. However, his condition worsened and despite all efforts, the patient experienced rapid deterioration and ultimately succumbed to disease progression.

Materials and methods

Hematoxylin and eosin (H&E)–stained sections of the thigh mass core biopsy were reviewed. Immunohistochemical stains performed at the referring institution were reviewed, and additional immunohistochemical stains were performed at the receiving institution using an automated immunohistochemistry platform (Benchmark XT; Ventana Medical Systems, Tucson, AZ, USA) on 4 µm formalin-fixed, paraffin-embedded (FFPE) tissue. The antibody results are reported below.

Interphase fluorescence in situ hybridization (FISH) using commercially available probes (Abbott Laboratories, Abbott Park, IL) was performed on FFPE tissue specimens following standard FISH pretreatment protocols and hybridization with dual-color break-apart (BAP) probes for MYC, BCL2, and BCL6 and a dual-color, dual-fusion (D-FISH) probe for MYC/IGH. A total of 100 interphase nuclei were independently evaluated for each probe set by two technologists (50 nuclei each) for each population and reported as a percentage of abnormal nuclei using CytoVision software (Leica, Buffalo Grove, IL).

Immunoglobulin (Ig) gene rearrangement studies were performed on FFPE tissue specimens, with each population differentially dissected and tested separately. DNA extraction was performed using standard procedures. Commercially available primers were used to amplify rearranged Ig heavy and Ig kappa light chain genes using polymerase chain reaction (PCR). The PCR products were analyzed by capillary electrophoresis.

Results

Histologic examination of the thigh mass showed skeletal muscle and soft tissue involved by sheets of atypical cells (Fig. 1c, d). Approximately three-quarters of the infiltrate was composed of intermediate to large lymphoid cells that stained positive for B-cell lineage markers including CD19 and CD20, and were negative for CD163 and CD68 (KP1) (Fig. 1c, e, g, i, k). This population was also positive for CD22, CD79a, and PAX5, and was negative for additional histiocytic markers including lysozyme, CD11c, and CD14. FISH analysis of this B-cell lineage population demonstrated MYC and BCL2 rearrangements with the same rearrangement pattern as the initial diagnostic lymph node biopsy, supporting involvement by the patient’s known double-hit lymphoma (Fig. 3a–d). The remaining infiltrate was morphologically distinct, composed of large, round to oval cells with irregular nuclear contours, vesicular chromatin, prominent nucleoli, and abundant, lightly eosinophilic cytoplasm (Fig. 1d). The biopsy was fragmented, and some fragments contained alternating areas of the two populations. This second population was negative for B-cell lineage markers CD19 and CD20, and was positive for CD163 and CD68 (KP1) (Fig. 1f, h, j, l). This population was also positive for other histiocytic markers including lysozyme, CD11c (partial), and CD14 (partial), and was negative for all other tested B-cell lineage markers including CD22, CD79a, and PAX5. BRAF V600E was negative in both components. Mitotic figures and apoptotic bodies were easily identified in both components and a high proliferation index of approximately 90% was observed. FISH analysis of this histiocytic population revealed the same MYC and BCL2 rearrangements as was detected in the B-cell lineage population (Fig. 3e, f). Ig gene rearrangement studies detected an identical gene rearrangement pattern in both the B-cell and histiocytic populations, further supporting the clonal relationship between the two populations (Fig. 4a, b).

Fig. 3
figure 3

Fluorescence in situ hybridization (FISH), pre- and post-CAR T-cell therapy. a, b FISH performed on left inguinal lymph node biopsy, pre-CAR T-cell therapy, showed MYC rearrangement (a), demonstrated by break-apart FISH probe (red arrows indicate separation with two copies of 5ʹMYC[8q24]). BCL2 rearrangement (b) is demonstrated by break-apart FISH probe (red and green arrows indicate separation of 5ʹBCL2[18q21] and 3ʹBCL2[18q21], respectively). c–f FISH performed on the right thigh mass biopsy, post-CAR T-cell therapy, showed MYC (c) and BCL2 (d) rearrangements (arrows) in the high-grade B-cell lymphoma component. MYC (e) and BCL2 (f) rearrangements (arrows) are also present in the histiocytic sarcoma/malignant histiocytosis component

Full size image
Fig. 4
figure 4

Immunoglobulin (Ig) gene rearrangement studies. The high-grade B-cell lymphoma component (a) and the histiocytic sarcoma/malignant histiocytosis component (b) show an identical gene rearrangement pattern

Full size image

Discussion

Lymphoma transdifferentiation and potential mechanisms

As more patients are treated with CAR T-cell therapy, great insights are gleaned into mechanisms of relapse, ranging from loss of CD19 expression to lineage switching of the malignant cells. To our knowledge, this is the first reported case of transdifferentiation of high-grade B-cell lymphoma with MYC and BCL2 rearrangements into histiocytic sarcoma/malignant histiocytosis following anti-CD19 CAR T-cell therapy. This represents a very aggressive mode of relapse, as illustrated by the patient’s rapid clinical decline.

Distinguishing transdifferentiation from de novo disease requires evidence that the two entities possess the same genetic alterations. In our case, core biopsies of the right thigh mass revealed two morphologically and immunophenotypically distinct cell populations; however, the presence of the same MYC and BCL2 rearrangements detected by FISH analysis and the same Ig gene rearrangement pattern in both the B-cell component and the histiocytic component supports that these two processes are clonally related. Secondary HS/MH can develop from transdifferentiation events, and it typically occurs metachronously with another hematologic malignancy. In a study of eight patients with follicular lymphoma (FL) and histiocytic/dendritic neoplasms, clonal relationships were established by the presence of t(14;18) in both components [8]. Similar to our case, three of the eight cases had synchronous tumors, of which two had both components within the same specimen.

Rare cases of transdifferentiation after CAR T-cell therapy have been reported in the literature, including one patient with diffuse large B-cell lymphoma (DLBCL) transdifferentiation into dendritic sarcoma [11] and another with mantle cell lymphoma (MCL) transdifferentiation into poorly differentiated sarcoma [12]. In the latter case, genetic analysis performed on specimens from the patient’s early MCL, late MCL, and subsequent poorly differentiated sarcoma demonstrated the same IGH gene rearrangement and extensive epigenetic reprogramming as the mechanism for transdifferentiation [12]. The authors speculated that CAR T-cell therapy may have contributed to this event by creating selective pressure, allowing the sarcoma cells to proliferate. In another case, treatment of chronic lymphocytic leukemia with CAR T-cell therapy was followed by relapse with clonally related plasmablastic lymphoma [13]. Another study analyzed tumor samples from nine patients with various B-cell lymphomas before and after CAR T-cell therapy [5]. In seven out of nine analyzed patients, the post-treatment tumors displayed loss of one or more B-cell markers. Two patients had complete loss of B-cell markers and the post-treatment tumors were phenotypically distinct. Further analysis of one patient revealed epigenetic modifications as a possible mechanism for transdifferentiation of a B-cell neoplasm into a T-cell neoplasm.

The concept of anti-lymphoma therapy exerting selective pressure on neoplastic cells has been previously established [14]. In acute leukemia, downregulation of CD19 expression appears to be one mechanism for gaining resistance to CAR T-cell therapy [5, 15]. However, in B-cell lymphomas, the frequency of CD19 downregulation associated with CAR T-cell therapy failure remains a controversial topic, as there are many proposed mechanisms of resistance. Some of them involve tumor factors which have been categorized as relapse with positive target antigen expression and relapse with negative target antigen expression [16]. In our case, it can be noted that the B-cell component post-CAR T-cell therapy retained target antigen expression, which certainly could have contributed to the poor outcome experienced by the patient, in addition to the transdifferentiation event.

HS/MH treatment approach and prognosis

HS/MH can involve multiple organs, including lymph nodes, skin, liver, spleen, lungs, bone, and central nervous system; therefore, the clinical presentation is variable. Most cases present as extranodal disease involving the skin, gastrointestinal tract, and soft tissues. The minority of cases (< 20%) present with solitary lymph node involvement [7, 12, 17, 18]. Systemic presentations showing disseminated MH have also been reported [18].

HS/MH has an aggressive clinical course, and it carries a poor prognosis due to clinically advanced disease presentation, rapid progression, and poor response to therapy [19,20,21]. From a prognostic standpoint, it is very important to distinguish secondary HS/MH from primary, as it has a significantly shorter overall survival (11.8 months compared to 70 months, respectively) [22].

Given the rarity of this neoplasm, there remains no standardized therapy. Like other malignancies, treatment options include surgery, radiotherapy, and systemic chemotherapy depending on the extent of disease. Most of the clinical experience to date is based on case reports and case series from various centers, which can provide limited general guidelines.

For localized disease, the preferred approach is resection with appropriate margins whenever possible. In a population-based analysis of 158 cases, patients without bone marrow, spleen, or reticuloendothelial system involvement who were managed surgically had a better OS than those who were not [7].

Systemic treatment

Historically, most cases of multifocal HS/MH have been treated with lymphoma-type regimens such as CHOP and ICE [19, 23], possibly because they were misdiagnosed as non-Hodgkin lymphomas [19]. Specific data about regimens and outcomes is scant in the literature and limited to small case series [18, 19, 23].

Monotherapy with vinblastine and/or prednisone [23] is an option for patients with poor performance status who are not candidates for combination regimens. It is also hypothesized that those who achieved remission after chemotherapy alone should be considered for allogeneic stem cell transplantation to prevent recurrence [24].

Patients with relapsed HS/MH and BRAF V600E mutations have been reported to have a response to vemurafenib [25]. Mutations in the MAP2K1 pathway may be amenable to MEK inhibitors such as trametinib [26]. Alemtuzumab has also been used successfully in some cases [27]. Lastly, increased expression of Programmed Cell Death 1 Ligand in HS and responses to PD-1 inhibitors have been described [28, 29].

This case illustrates the diagnostic and therapeutic challenges associated with HS/MH, further complicated by a rare transdifferentiation event after CAR T-cell therapy. From a diagnostic perspective, transdifferentiation is uncommon but should be considered when a suspected recurrent neoplasm diverges from the original diagnosis. Pursuing additional lineage markers is critical in these situations.

From a therapeutic perspective, cases such as these are exceptionally difficult because a treatment standard does not exist. As stated above, current sparse data suggests that HS/MH is often resistant to chemotherapy. The optimal therapy can only be ascertained in a prospective study. Given the rarity of this disease, sizeable well-designed clinical trials pose numerous difficulties. Ultimately, patients should be encouraged to enroll in a clinical trial whenever possible. We are on track to understanding cell reprogramming that may give rise to new therapeutic targets to improve survival in patients with this aggressive neoplasm.