Spatial Phenotyping of Nodular Lymphocyte Predominant Hodgkin Lymphoma and T-cell/Histiocyte-Rich Large B-cell Lymphoma

doi:10.21203/rs.3.rs-3629195/v1

Download PDF

Article

Spatial Phenotyping of Nodular Lymphocyte Predominant Hodgkin Lymphoma and T-cell/Histiocyte-Rich Large B-cell Lymphoma

https://doi.org/10.21203/rs.3.rs-3629195/v1

This work is licensed under a CC BY 4.0 License

Journal Publication

published 31 May, 2024

Read the published version in Blood Cancer Journal →

You are reading this latest preprint version

Nodular lymphocyte-predominant Hodgkin lymphoma (NLPHL) is a rare lymphoma with sparse tumor B-cells and a favorable prognosis. Variant growth patterns of NLPHL, however, show advanced stage, progression to T-cell/histiocyte-rich large B-cell lymphoma (THRLBCL) and a worse prognosis. We studied the spatial configuration of the tumor microenvironment (TME) of NLPHL and THRLBCL using highplex imaging to capture single-cell parameters including spatial localization in 20 patient samples of NLPHL and THRLBCL. Our findings show distinct spatial configurations and TME composition that differ among typical and variant NLPHL, and THRLBCL. Tumor B-cell size and content was lowest in typical NLPHL, followed by variant NLPHL, and highest in THRLBCL, whereas an opposite trend characterized TME B-cells. Typical NLPHL showed abundant helper T-cell subsets, while THRLBCL showed abundant cytotoxic T-cells and monocytes. Spatial analysis further revealed specific interactions typical of NLPHL patterns and THRLBCL. CD4/CD8 double-positive T-cells were detected in all NLPHL but not in the majority of THRLBCL, and were found to be spatially distant from tumor B-cells and TFH-rosettes. We conclude that our results provide valuable insights into immunoarchitectural configurations that inform differences in biologic behavior and could aid in the development of future therapeutics for patients affected by this spectrum of lymphomas.

Nodular lymphocyte predominant Hodgkin lymphoma

T-cell/histiocyte-rich large B-cell lymphoma

PhenoCycler

CODEX

multiplex imaging

The tumor microenvironments in NLPHL and THRLBCL show distinct spatial organization and are not randomly distributed.
Tumor cell size, tumor-associated T-cell and monocyte subsets, and their spatial interactions differ among typical and variant NLPHL, and THRLBCL.
CD4/CD8 double-positive T-cells are seen in all NLPHL but not in the majority of THRLBCL and are spatially distant from LP-cells and TFH-rosettes.

Nodular lymphocyte predominant Hodgkin lymphoma (NLPHL; also called nodular lymphocyte predominant B-cell lymphoma) is a rare neoplasm derived from germinal center (GC) B-cells.^[1–4] Although the prognosis of NLPHL is favorable with a 10-year overall survival of > 80%, up to 20–30% of patients experience recurrences and/or progression to large B-cell lymphoma, particularly, T-cell/histiocyte-rich large B-cell lymphoma (THRLBCL).^[3–14] The neoplastic lymphocyte-predominant (LP) cells of NLPHL form a small fraction of the tumor mass and are embedded in a rich tumor microenvironment (TME). We previously described six immunoarchitectural growth patterns of NLPHL which were found to correlate with disease recurrence and progression.^[15] The prognostic relevance among the two “typical” and four “variant” growth patterns was subsequently confirmed and validated in independent adult and pediatric patient cohorts.^{[6, 10, 16–18]} The recognition of variant growth patterns and their relationship to prognosis has prompted inquiries about the biologic basis that underlie these histologically-defined patterns, but has remained largely enigmatic.

De novo THRLBCL, unlike NLPHL, is associated with an aggressive clinical course, although an indolent course is seen in cases that arise from NLPHL.^{[3, 6, 19–20]} NLPHL and THRLBCL are increasingly becoming recognized as a true biologic continuum due to striking similarities at the phenotypic and genetic levels. Despite extensive gene expression and mutational profiling studies, reproducible criteria to distinguish the variant patterns of NLPHL, particularly pattern E, from THRLBCL has yet to be established.^{[6, 17, 21–22]} Therefore, we studied the TME composition and spatial organization at a single cell level using CO-Detection by indEXing [CODEX® (rebranded as PhenoCycler™) ^[23–26] and a curated 21-antibody panel that was specifically designed to target tumor B-cells and immune cell subsets in the TME. By doing so, we gained insights into previously unexplored cellular interactions that could shed light on different growth characteristics associated with the biological behavior of NLPHL and THRLBCL.

Tissue samples: Formalin-fixed paraffin-embedded (FFPE) tissue from 20 lymph node excision biopsies of 16 NLPHL (8 typical and 8 variant growth patterns) and 4 THRLBCL were retrieved from the Stanford Pathology archive with Institutional Review Board approval. The cases were reviewed by two pathologists (SY and YN). For analysis purposes, NLPHL cases were categorized into 3 groups according to Fan et al^[15]: group 1, typical nodular growth patterns A and B; group 2, variant nodular growth patterns C and D; group 3, variant diffuse growth pattern E. None had diffuse pattern F. THRLBCL cases comprised group 4 (Table 1, Fig. 1).

Table 1

Diagnoses with variant growth patterns and group designations for NLPHL
CODEX ID	Diagnosis	Subtype	Proportion of Variant Patterns	Designation for Current Study	Age	Gender	Stage	First presentation site (LN)	Prognosis
NC1	NLPHL	Typical	NA	Group 1	53	Male	2	Inguinal	Recurrence, alive
NC2	NLPHL	Typical	NA	Group 1	38	Female	1	Mesenteric	Not relapsed, alive
NC3	NLPHL	Typical	NA	Group 1	22	Female	2	Cervical and axillary	Not relapsed, alive
NC4	NLPHL	Typical	NA	Group 1	50	Male	2	Cervical/supraclavicular/occipital/pre-auricular	Not relapsed, alive
NC5	NLPHL	Typical	NA	Group 1	44	Male	3	Mediastinal	Not relapsed, alive
NC6	NLPHL	Typical	NA	Group 1	45	Male	4	Axillary	History of THRLBCL, not relapsed, alive
NC7	NLPHL	Typical	NA	Group 1	22	Male	2	Cervical	Not relapsed, alive
NC8	NLPHL	Typical	NA	Group 1	9	Male	3	Cervical	Recurrence, alive
NV1	NLPHL	Variant	Variant C (70%)	Group 2	7	Male	2	Cervical	Not relapsed, alive
NV3	NLPHL	Variant	Variant C (70%)	Group 2	28	Male	3	Submandibular LN	Not relapsed, alive
NV5	NLPHL	Variant	Variants C (70%), D (10%)	Group 2	45	Male	2	Mesenteric LN	Not relapsed, alive
NV6	NLPHL	Variant	Variant C (30%)	Group 2	69	Female	2	Paratracheal LN	progressed to DLBCL, alive
NV7	NLPHL	Variant	Variant D (20%)	Group 2	67	Male	4	Cervical	Not relapsed, alive
NV4	NLPHL	Variant	Variant E (80%)	Group 3	67	Male	4	Cervical	NA
NV9	NLPHL	Variant	Variants E (60%), C (10%), D (10%)	Group 3	15	Male	2	Inguinal	Not relapsed, alive
NV10	NLPHL	Variant	Variant E (30%)	Group 3		Male	2	Cervical	Not relapsed, alive
TD1	THRLBCL	NA	NA	NA	21	Female	4	Abdominal	NA
TD5	THRLBCL	NA	NA	NA	36	Male	1	Submandibular	NA
TD6	THRLBCL	NA	NA	NA	25	Male	4	Axillary LN	Died
TD7	THRLBCL	NA	NA	NA	19	Male	4	Axillary LN	NA
*IGD + LP cells

CODEX antibody conjugation and validation: 50ug of carrier-free antibodies (Table S1, Figure S1), were conjugated to DNA oligonucleotide barcodes using conjugation kits (Akoya Biosciences, CA). Antibodies were concentrated on a 50kDa filter equilibrated with filtration buffer and sulfhydryl groups activated by incubating for 30 minutes at room temperature (RT) with reduction mix. Barcodes added to antibodies were incubated for 2hrs at RT, eluted in 100ul storage buffer, spun at 3000g for 4 minutes and stored at 4⁰C until use. Conjugated antibodies were validated individually on positive control tissue and/or cell line pellets prior to the multiplex run. Fluorescent probes complementary to the DNA barcodes were added to the stained tissue for 5 minutes and imaged.

CODEX antibody staining: A 15x15mm representative area was selected from 4µm FFPE sections and mounted on CODEX^R coverslips for staining with a 21-antibody cocktail. Coverslips were deparaffinized, rehydrated, and antigen retrieval performed using a pressure cooker for 20 minutes in 1X citrate buffer, pH 6.0 (NovusBio) followed by heating for 20 minutes at high pressure in 1X EDTA, pH 9.0 (Abcam). Autofluorescence was bleached as per protocol.^[33] The antibody cocktail was added to coverslips and stained in a sealed humidity chamber at 4^oC overnight, washed, hydrated, and fixed according to manufacturer’s protocol.

CODEX multicycle setup and imaging: Coverslips were mounted onto Akoya’s custom made stage and stained with Hoechst nuclear stain at a 1:2000 dilution in 1X CODEX buffer (Akoya). A 96-well plate was used to set up the multicycle experiment with different fluorescent oligonucleotides. Fluorescent oligonucleotides were added to a reporter stock solution (1:150 Hoechst stain and 1:12 dilution of assay reagent in 1X CODEX buffer) at a final concentration of 1:50 in a total of 250 ul per well. A blank cycle without fluorescent probes was performed at the start and end of the experiment to capture autofluorescence.

Automated image acquisition and fluidics was performed using Akoya’s software driver CODEX Instrument Manager (CIM, version 1.29) and the CODEX platform. Imaging was performed using a Keyence BZ-X810 microscope, fitted with a Nikon CFI Plan Apo 20X/0.75 objective. 11 z steps were acquired with the pitch set at 1.5 in the BZ-X software. Raw tiff files were processed using the CODEX Processor version 1.7.0.6 by Akoya to generate a large, stitched tiff file. Data processing included deconvolution, background subtraction and drift correction. Segmentation was performed using the nuclear signal by finding a local intensity maxima and then defining the radius for growth around this maxima. Threshold values for region growth are user-defined, and we used a radius of 5 and a size cut off factor of 0.1 to define the volume of a cell for accurate segmentation results. Segmentation accuracy was morphologically checked across ROIs by pathologists as shown in S5. The segmentation algorithm accounts for spillover compensation of signal spillover from neighboring cells. Segmentation masks, spatial coordinates and marker intensities for all detected cells were then exported for further analysis (Figure S2).

CODEX analysis: Four representative regions of interest (ROIs) were digitally selected from each case by two pathologists to ensure NLPHL patterns were adequately represented (Figure S3). The selected ROIs were comparable across all 20 NLPHL and THRLBCL cases and the number of cells per ROI ranged between 10431–65884, mean+/-SD of 33965.7 +/- 14331.6. Multiplex Analysis Viewer (MAV, Akoya), an ImageJ plugin, was used to visualize and validate marker expression, determine segmentation accuracy, extract ROI’s and gate data for further analysis.

Clustering was performed in R using the Seurat package for unsupervised clustering ^[27–32]. Clustering identified sub-populations according to the expression of different markers within the DAPI-positive population. The resulting clusters were overlaid on the fluorescent images based on X/Y positions and visually verified based on marker expression, morphology, and localization within the tissue. Similar clusters were manually merged and annotated. Populations were defined based on coexpression of defining markers and confirmed by absence of other non-specific markers to diminish the effect of overlapping signals. Post-clustering annotation of cellular phenotypes was performed by the two pathologists.

Spatial analysis was performed using MAV, which defines interactions based on spatial proximity. We used the distance of 3–30 µm to define the distance used to calculate the number of interacting cells. An interacting cell as defined by MAV is any cell that falls within the predefined spatial distance range. To better define TME cells for study, we broadly grouped cell populations by phenotype and by their activation status as defined in Table 2. All markers showed standard performance in both settings except ICOS, IDO1, MUM1 and PDL1, which showed inconsistency among cases and were selectively excluded at the analysis level.

Table 2

Z-scores of cellular populations analyzed in NLPHL groups and THRLBCL
Cell Types (Z-Scores)	Phenotype	NLPHL			THRLBCL
Cell Types (Z-Scores)	Phenotype	Group 1	Group 2	Group 3	THRLBCL
Tumor B *	CD20/BCL6	-0.47	-0.35	0.36	1.10
TME B	CD20/IgD	0.77	0.28	-0.38	-1.61
Helper T	CD3/CD4	0.50	0.15	-1.06	-0.40
Activated Helper T	CD3/CD4/LAG3/CD69	-0.14	-0.00	1.01	-0.48
Follicular Helper T (TFH)	CD3/CD4/PD1/BCL6	0.16	0.32	-0.05	-0.68
Activated TFH (aTFH)	CD3/CD4/PD1/BCL6/LAG3/CD69	-0.55	-0.37	0.52	1.18
Cytotoxic T-cells	CD3/CD8	-0.05	-0.04	-0.21	0.29
Activated cytotoxic T	CD3/CD8/LAG3/CD69	-0.39	0.19	0.03	0.50
Regulatory T-cells (T-reg)	CD3/CD4/FOXP3	-0.31	0.14	0.12	0.34
CD4/CD8 double-positive T (DPT)	CD3/CD4/CD8	0.17	0.24	-0.09	-0.59
Monocytes	CD68	-0.21	-0.29	-0.03	1.02
Immature Monocytes	CD68/CD14	-0.54	0.12	0.18	0.79
* Lymphocyte Predominant (LP)-cells in NLPHL and large, atypical B-cells in THRLBCL

Tumor cell size analysis: To assess tumor cell size, MEF2B immunohistochemistry (polyclonal, 1:200 dilution; Sigma-Aldrich, St Louis, MO), which targets tumor cell nuclei only, was employed. We utilized Qupath (V0.4.4; https://qupath.github.io/) to quantify nuclear area and perimeter of MEF2B positive tumor cells.

Statistical analysis: R version 4.1.2 in R server and R console environments were used to quantify cell population abundance per region/case. To consolidate values across regions within each case, means were computed. For patterns, values were averaged across regions that matched the pattern. Significance testing at the region level was performed using rank-sum wilcoxon t-tests. The abundances were normalized to mean zero and unit variance. For visualization, R packages 'ggplot2' for bar and box plots and 'ComplexHeatmap' for heatmaps were used. To explore interactions among cell populations, cell counts and interaction matrices describing each population within each region from MAV were extracted. To account for differences in the cell population sizes, we corrected each interaction by dividing by the total number of cells involved in the interaction. We conducted rank-sum Wilcoxon t-tests for testing the significance of these interactions. For visualizing the interactions, we used the 'chordDiagram' R package for Fig. 3 and the 'corrplot' R package for Fig. 4. All P-values were two-tailed and considered significant at P < 0.05.

Study populations in NLPHL and THRLBCL: The 16 NLPHL evaluated in 3 analysis groups, and 4 THRLBCL are shown (Table 1, Fig. 1). By utilizing a multiplex panel of 21 markers, tumor B-cells, TME B-cells, T-cells, and monocyte subsets, inclusive of their functional status, were defined and visualized (Fig. 2, Figure S1). Markers validated individulally and by multiplex runs were visualized as defined cell populations in every case (Fig. 2, Tables 2), and in each of the ROIs selected (Figure S3). Z-scores allowed comparison of the proportion of each cellular phenotype in relation to the overall number of cells within each ROI (Tables 2 and S4, and Fig. 2, S4-5), and spatial intractions were calculated as shown (Tables 3, 4, S5 and Figs. 3–5).

Table 3

Interaction counts between LP cells and TME components
Cell Type	NLPHL			THRLBCL
Cell Type	Group 1	Group 2	Group 3	THRLBCL
TFH	0.094	0.058	0.079	0.075
Regulatory T-cells	0.029	0.029	0.029	0.027
Tumor B-cells	0.26	0.24	0.29	0.51
B-cells	0.065	0.062	0.047	NA
Monocytes	0.039	0.023	0.039	0.05
CD4/CD8 T-cells	0.027	0.034	0.034	NA
Cytotoxic T-cells	0.046	0.034	0.044	0.04
Helper T-cells	0.056	0.045	0.047	0.095

Tumor B-cell and TME B-cell abundance vary across NLPHL and THRLBCL: THRLBCL had a relatively higher number of tumor B-cells followed by groups 3, 2 and 1 NLPHL (groups 3 vs 1, p = 0.009; and groups 3 vs 2, p = 0.05). In contrast, TME B-cells in NLPHL showed significantly higher TME B-cells in groups 1 vs 2 and 3 (p = 0.003 and < 0.0001 respectively). The TME B-cells in THRLBCL were so sparse that it could not be detected using the clustering algorithm.

T-cell abundance and composition vary across NLPHL and THRLBCL: T-helper (TH) cells were significantly more abundant in groups 1 and 2, and relatively less predominant in group 3 and THRLBCL. The helper T-cell composition was strikingly higher in groups 1 vs. 3 and THRLBCL (p = 0.04 and p = 0.025 respectively), with relatively higher number of TH cells in THRLBCL compared to group 3. No significant difference between helper T-cell counts were detected in groups 1 and 2.

Activated TH cells were notably abundant in all NLPHL groups compared to THRLBCL, and showed the highest abundance in group 3. There were also a significantly higher number of activated TH cells in group 3 vs. THRLBCL [p = 0.01]. There was an abundance of helper and activated TH cells between group 3 and THRLBCL; TH cells were more abundant in THRLBCL and their activated form was predominant in group 3.

TFH cells were more abundant in all groups of NLPHL relative to THRLBCL; however, this observation did not reach statistical significance. Interestingly, there were a number of statistically significant findings involving activated TFH cells: there was significantly lower abundance of activated TFH cells in group 1 in contrast to group 3 and THRLBCL [p = 0.001 and < 0.0001 respectively]. Similarly, there was a lower adundance of activated TFH cells in groups 2 vs 3 and THRLBCL (p = 0.03 and 0.002 respectively). Overall, activated TFH cell abundance was reverse of the abundance of TH cells across NLPHL groups and THRLBCL. T-regs were deteced in abundance in groups 2 and 3 as well as THRLBCL, but were significantly less abundant in group 1 (groups 3 vs 1, p = 0.004).

Both cytotoxic and activated cytotoxic T-cells were abundant in THRLBCL relative to all groups of NLPHL. Z-scores for cytotoxic T-cells were highest in THRLBCL, followed by group 1 and 2, and were the lowest in group 3. This observation did not show statistical significance. Activated cytotoxic T-cells were most abundant in THRLBCL, low in groups 2 and 3, and least abundant in group 1 (THRLBCL and group 1, p = 0.04).

Additionally, CD4/CD8 double positive T (DPT)-cells were detected in a substantial proportion of NLPHL and included group 1 (7 of 8, 87.5%), group 2 (4 of 5, 80%), and group 3 (2 of 3, 66.7%). Among THRLBCL, only 1 of 4 (25%) cases showed CD4/CD8 DPT-cells.

Monocytes are increased in THRLBCL compared to NLPHL: Both mature and immature monocytes were more adundant in THRLBCL compared to all NLPHL groups. In addition, THRLBCL showed a statistically significant abundance in monocyte count in comparison to group 2 (P = 0.036). There was differential abundance of immature monocytes, which were most abundant in THRLBCL, followed by groups 3, 2 and 1 in descending order of abundance (THRLBCL vs. groups 1 and 2, p = < 0.001 and 0.017 respectively). Although group 3 showed a lower abundance of immature monocytes compared to THRLBCL, the difference was not statistically significant

Tumor B-cells show differences in spatial localization in NLPHL and THRLBCL: The cells with the strongest interaction/spatial proximity to LP cells in group 1 were TFH cells followed by TME B-cells, whereas cytotoxic and TH cells were not found to interact strongly with LP cells. In group 2, a similar spatial proximity was found and included TFH cells and TME B-cells; however, TME B-cells were found to be more proximal to LP cells than TFH cells. Furthermore, there was a statistically significant stronger interaction between LP cells and TFH cells in group 1 compared to group 2 (p = 0.04). In group 3, TFH cells showed the most spatial proximity to LP cells, followed by TH cells and TME B-cells. Monocytes and cytotoxic T-cells were spatially distant from LP cells. Finally, in THRLBCL, TH cells were found to show the highest proximity to tumor B-cells, followed by TFH cells, and then monocytes and cytotoxic T-cells.

TME B-cells are spatially proximal to helper and regulatory T-cells in typical NLPHL: A higher level of spatial interaction was found between TME B-cells and both helper T-cell and T-regs in group 1 NLPHL. The helper T-cell interaction with TME B-cells in group 1 was significantly stronger compared to groups 2 and 3 (p = 0.002, 0.035 respectively). Similarly, T-regs interacted strongly with TME B-cells in group 1 compared to groups 2 and 3 (p = 0.002, 0.013 respectively). In group 1 =, although TME B-cells follow TFH cells in their interaction with LP cells, TME B-cells also interact significantly with TH cells and T-regs.

Helper T-cell interactions show spatial variation in NLPHL and THRLBCL: TH and cytotoxic T cell interactions showed significant differences in nodular versus diffuse configurations across NLPHL and THRLBCL. There was a notable, significantly stronger interaction between helper and cytotoxic T-cells in group 1 compared to group 3 and THRLBCL (p = 0.002, and 0.017 respectively). Similarly, the interaction between both helper and cytotoxic T-cells was more pronounced in group 2 compared to group 3 and THRLBCL (p = 0.012 and 0.03 respectively). TH cells showed strongest likelihood of interaction with monocytes in THRLBCL and group 1. There was no significant difference between both groups regarding this interaction, however, helper T-cell and monocyte interaction was significantly higher in group 1 compared to groups 2 and 3 (p = 0.06, < 0.0001 respectively), and between group 2 and group 3 NLPHL (p = 0.012).

CD4/CD8 DPT-cells are spatially removed from LP cells: CD4/CD8 DPT-cells were spatially distant from the vicinity of LP cells and TFH-rosettes where most TME B-cell interactions with LP cells were detected. The DPT-cells showed stronger interaction with cytotoxic T-cells in groups 1 vs 2 (p = 0.047). These DPT-cells also interacted more with monocytes and T-regs in group 1 compared to group 3 (p = 0.02, 0.03 respectively). Overall, cytotoxic T-cells, T-regs, CD4/CD8 DPT-cells and monocytes were spatially distant from LP cells when compared with TME B-cells and TFH cells.

Cytotoxic T-cell interactions with T-regs and monocytes show spatial variation: Cytotoxic T-cells

exhibit closer proximity to monocytes in THRLBCL in comparison to groups 1, 2, and 3 (p > 0.05,

0.019, 0.002, respectively). Furthermore, cytotoxic T-cells and TFH cells were found to have the

strongest spatial relationship in groups 2 and 3 as well as THRLBCL compared to group 1

(p = 0.013, > 0.05, and 0.019, respectively). This relationship was strongest in group 2 NLPHL compared to all other cases and was least consequential in THRLBCL. Moreover, cytotoxic T-cells and T-regs showed a significant spatial proximity in groups 1 and 2 compared to THRLBCL (p = 0.0006, 0.006 respectively).

Monocyte interactions in group 1 NLPHL and THRLBCL: Spatial proximity between monocytes and T-regs was prominent in group 1 NLPHL compared to group 3 and THRLBCL (p = 0.003, 0.048 respectively). In contrast, the proximity between monocytes and TFH cells was the highest in THRLBCL. TFH-monocyte interaction was highest in THRLBCL compared to groups 1, 2 and 3 NLPHL (p = < 0.0001, 0.0001, 0.001 respectively). Monocyte interactions with LP cells were less prominent in group 2 compared to groups 1 and 3 NLPHL [p = 0.04, 0.035 respectively].

T-reg interactions are unique to group 1 NLPHL: T-regs showed strong interactions with TME B-cells, monocytes, cytotoxic T-cells, and CD4/CD8 DPT-cells in group 1 NLPHL. In contrast, T-reg and TFH populations showed stronger interactions that were statistically significant in groups 2 and 3, as well as THRLBCL in comparison to group 1 NLPHL (p = 0.004, 0.002, and 0.02 respectively). The overall abundance of T-regs was, however, the lowest in group 1 among all cases evaluated. Our spatial analysis data takes into consideration the number of interactions between populations in relation to the number of cells in each of the studied groups.

NLPHL and THRLBCL tumor cell size measurements: Tumor cell nuclear area and perimeter showed a gradual increase from group 1 to 2 to 3 NLPHL. These findings were statistically significant (p < 0.001 for group 3 vs 2, group 3 vs 1 and group 2 vs 1, p < 0.001 with regard to nuclear area. Similar statistical significances were noted with nuclear perimeter (p < 0.001 for group 3 vs 1 and group 3 vs 1 NLPHL; and p = 0.018 for group 3 vs 2 NLPHL). Both nuclear size and perimeter were smaller in THRLBCL cases compared to group 3 NLPHL (p < 0.001 for THRLBCL vs group 3 for both nuclear area and perimeter). However, nuclear size of THRLBCL remained higher than both groups 1 and 2 NLPHL (p < 0.001 for THRLBCL vs groups 1 and 2 NLPHL for both nuclear area and perimeter) (Figure S6).

NLPHL and THRLBCL are considered to represent different endpoints of the same disease spectrum.^[6] Gene expression profiling, copy number alterations, and mutational profiling studies have further underscored the extensive overlap between these entities.^{[6, 33–41]} Although the drivers of transformation from NLPHL to THRLBCL remain largely unknown, the differences in prognosis highlight the critical need for more robust criteria to distinguish NLPHL, especially pattern E, from THRLBCL. Our results provide valuable insights to understand tumor-TME dynamics and their clinical implications.

Both NLPHL and THRLBCL have sparse tumor B-cells; however, our data clearly showed an ordered abundance of tumor B-cells from least abundant in typical patterns, followed by variant patterns of NLPHL, and most abundant in THRLBCL. The cellular composition of the TME surrounding LP cells is postulated to result from accumulation of genomic aberrations due to neo-antigen formation by LP cells. ^{[39, 39–42]} Our data captured substantial differences both proximal to and distant from tumor B-cells. Spatial analysis showed that TFH cells were the closest to LP cells in typical NLPHL and confirm prior observations that TFH-rosettes were most abundant in typical NLPHL,^{[41, 43–44]} less frequent in variant patterns, and rare in THRLBCL.^{[6, 17, 45–46]} In addition, we demonstrate for the first time that there is a gradual increase in tumor cell size from the smallest tumor cells in typical NLPHL, followed by variant NLPHL, and the largest tumor cells in THRLBCL. This graded increase in tumor B-cell size is aligned with the presence or absence of nodularity in the overall immunoarchitecture of NLPHL and THRLBCL. Nodularity is defined by the presence of follicular dendritic cell (FDC) meshworks, which typically show disruption or loss in variant patterns and absence in pattern E and THRLBCL. Although nodularity and FDC meshworks were not formally evaluated in this study, their influence on tumor cell size may have profound implications for lymphoma progression and capacity for dissemination in the less confined architectural configurations that are encountered in variant NLPHL and THRLBCL in contrast to typical NLPHL.

Spatial analysis further revealed insights into TME composition that could potentially differentiate NLPHL from THRLBCL, explain clinical behavior, and guide therapy. Notably, TFH abundance and activation status may influence TME dynamics since TFH cells were abundant in all NLPHL groups compared to THRLBCL, while activated TFH cells were more abundant in diffuse compared to nodular NLPHL. TFH cells showed more interaction with LP cells of NLPHL than tumor cells of THRLBCL; their presence was found to confound diagnosis when TFH-derived lymphomas were in the differential diagnosis.^[47] Also, TFH cells strongly interact with both monocytes and cytotoxic T-cells in THRLBCL and highlight the importance of an activated TFH component that differs among NLPHL patterns and THRLBCL.

Our data yielded three unique TME features that distinguished NLPHL from THRLBCL: at least a 2-fold increase in activated TH and TFH cells in all NLPHL groups compared to THRLBCL, a remarkable abundance of monocytes and cytotoxic T-cells in THRLBCL relative to NLPHL, and detection of CD4/CD8 DPT-cells in NLPHL but not in the majority of THRLBCL.

Higher cytotoxic and activated cytotoxic T-cells content in THRLBCL compared to NLPHL of all patterns, particularly pattern E, suggests a potential role for activation status of cytotoxic T-cells in orchestrating architectural dismantling in the transition from NLPHL to THRLBCL. Our data also confirm high CD8 content in THRLBCL,^{[17, 48]} and higher cytotoxic T-cell/monocyte interactions in THRLBCL compared with NLPHL,^[49] although the functional status of cytotoxic T-cells in driving NLPHL transformation remains unknown. In addition, we show that variant NLPHL patterns, especially pattern E, has stronger cytotoxic T-cell interactions with LP cells compared to typical nodular patterns of NLPHL. This finding may have therapeutic implications given that monocyte/cytotoxic T-cell interactions have been shown to influence responses to immune therapy and prognosis. ^[50]

A significantly higher number of immature monocytes detected in THRLBCL in our cohort, raises the possible acquisition of a tolerogenic TME, which has been shown to correlate with suppression of T-cell proliferation, immune evasion, and tumor dissemination.^{[17, 48, 51]} The molecular signature of THRLBCL shows upregulation of interferon-dependent pathway genes which promote immune tolerance in lymphoma.^[52] In NLPHL, high macrophage content correlates with poor prognostic factors including a lower complete remission rate,^[53] bulky disease, and variant growth patterns.^[54] Our findings confirm prior observations regarding the abundance of macrophages in THRLBCL and variant pattern NLPHL. In addition, our data shows that immature monocytes form the bulk of this difference in macrophage content. Furthermore, we found that spatially, monocytes were located relatively closer to LP/tumor B-cells in NLPHL pattern E and THRLBCL compared to nodular patterns of NLPHL. In addition, monocytes in THRLBCL interacted strongly with both TFH and cytotoxic T-cells, which may influence response to therapeutic approaches targeting immune checkpoints as shown in other lymphomas with an inflamed TME. ^{[50, 55]}

Nodular NLPHL patterns had increased TME B-cells and TH cells but decreased activated TFH cells compared to NLPHL pattern E and THRLBCL. A strong interaction between TH cells and cytotoxic T-cells was seen in the nodular groups, whereas the interaction between TH cells and tumor cells was more pronounced in diffuse cases. To the best of our knowledge, this is the first description of the spatial configuration of TH and activated TH cells in NLPHL. The abundance and functional status of both TH cells and monocytes may not only influence lymphomagenesis through the regulation of chemokines and other mediators, but may also play a pivotel role in remodeling the TME during transformation. CD4 + T cells are known to produce pro-inflammatory cytokines that lead to the recruitment and activation of innate effector cells, such as macrophages and cytotoxic T- cells.^[4] The observed strong interactions between TH cells and both cytotoxic T-cells and monocytes, may contribute to the superior prognosis of typical NLPHL. On the contrary, the abundance of PD1-positive T cells in CHL is associated with a poor prognosis.^[56]

Distinction between pattern E NLPHL and THRLBCL could be aided by certain histologic features we discovered, especially in small biopsies. In pattern E NLPHL, TME-B cells, activated TH cells, TFH cells, and CD4/CD8 DPT-cells were increased in comparison to THRLBCL. In contrast, the numbers of monocytes and cytotoxic T-cells were significantly higher in THRLBCL as well as monocyte interactions with TH cells, cytotoxic T-cells, and TFH.

Typical NLPHL (group 1) showed several exclusive spatial interactions characterized by strong interaction between LP and TFH cells but weak interactions with cytotoxic T-cells, T-regs, and monocytes. Away from LP cells, T-regs, TH cells and CD4/CD8 DPT-cells exhibited specific interactions. Although the number of T-regs was the lowest in group 1 NLPHL, they showed specific interactions compared to the other groups. In contrast, the interaction between T-regs and TFH cells were the least in group 1 NLPHL. The number of T-regs was reported to be decreased in NLPHL compared to reactive follicular hyperplasia ^[57]. To the best of our knowledge, our study is the first to investigate the variation in T-reg composition and spatial interactions among NLPHL variants and THRLBCL, although their functional contribution awaits further study. There were no significant differences in nodular NLPHL (patterns A, B, C and D), in activated cytotoxic T cells, immature monocytes, or T-regs; however, the distinct cellular interactions noted exclusively in group 1 NLPHL above indicate that it is not the abundance of TMA components but rather the interactions among cell populations that is likely to underlie the histologic and associated prognostic differences.

Our results detected CD4/CD8 DPT-cells by unsupervised clustering in 67–77% of NLPHL with no difference between typical or variant patterns. These DPT-cells were also detected in a prior flow cytometry study of NLPHL.^[58] The DPT-cells showed no discernible interaction with LP cells and were spatially distant from LP cells and TFH rosettes. In group 1 NLPHL only, DPT-cell interactions with cytotoxic T-cells, monocytes, and T-regs were also detected. Given that only a single case of THRLBCL had DPT-cells, further investigation of a larger cohort of THRLBCL is warranted to confirm these observations.

The alignment of TH and TFH cell activation as revealed by increased PD1 and CD69 expression in diffuse pattern E and THRLBCL in our cohort is of interest. CD69, a marker of T-cell activation,^[33] has also been detected by flow cytometry in NLPHL.^[59] Upon activation, PD1 and CD69 are simultaneously expressed in activated TFH cells and induce an immunosuppressive microenvironment by induction of T-cell exhaustion,^[60] although a direct impact on tumor aggressiveness has yet to be elucidated.

In conclusion, by deeply profiling the cellular composition and spatial distribution of NLPHL and THRLBCL, we corroborate known associations and discover previously unrecognized novel cellular interactions. These data provide insights into the functional status of cell populations and their interactions in the TME milieu by combining the ability to detect specific expression patterns of individual markers in tissue, obtain quantitative data at a single-cell level, and profile spatial localization of tumor and TME populations. The ability to use whole tissue sections further provides the advantage of careful selection and annotation by pathologists of specific case types and regions of interest to investigate known and novel spatial phenotypes.

Data Sharing: The datasets generated and analyzed during the current study are available from the corresponding author on reasonable request.

Author contributions:

Designed the study: SY, YN

Performed the study: SY, AS, AK, SZ, MB, YN

Analyzed the data: SY, AS, MB, YN

Wrote the Paper: SY, AJ, MB, YN

Reviewed and edited the paper: SY, AS, AK, SZ, MB, YN

Disclosure of conflicts of interest:

S.Y. reports research funding from Kite Pharma and California Institute for Regenerative Medicine; Y.N. reports consultancy from Leica Biosystems and Roche Diagnostics, and research funding from Kite Pharma; AK performed the entire study while at Stanford. Author recently joined Akoya Biosciences team.

Morton LM, Wang SS, Devesa SS. Lymphoma incidence patterns by WHO subtype in the United States, 1992–2001. Blood. 2006; 107: 265–276
Teras LR, DeSantis CE, Cerhan JR. US lymphoid malignancy statistics by World Health Organization subtypes. CA Cancer J. Clin. 2016; 66: 443–459
Alaggio R, Amador C, Anagnostopoulos I et al. The 5th edition of the World Health Organization Classification of Haematolymphoid Tumours: Lymphoid Neoplasms. Leukemia. 2022 Jul;36(7):1720-1748. PMID: 35732829.
Al-Mansour M, Connors JM, Gascoyne RD. Transformation to aggressive lymphoma in nodular lymphocyte-predominant Hodgkin’s lymphoma. J Clin Oncol. 2010; 28(5):793–799.
Biasoli I, Stamatoullas A, Meignin V. Nodular, lymphocyte-predominant Hodgkin lymphoma: a long-term study and analysis of transformation to diffuse large B-cell lymphoma in a cohort of 164 patients from the Adult Lymphoma Study Group. Cancer. 2010; 116(3):631–639.
Hartmann S, Döring C, Jakobus C. Nodular Lymphocyte Predominant Hodgkin Lymphoma and T Cell/Histiocyte Rich Large B Cell Lymphoma—Endpoints of a Spectrum of One Disease? PLoS ONE. 2013; 8, e78812.
Eyre TA, Gatter K, Collins, G. Incidence, management, and outcome of high-grade transformation of nodular lymphocyte predominant Hodgkin lymphoma: Long-term outcomes from a 30-year experience. Am. J. Hematol. 2015; 90: E103–E110.
Eichenauer DA, Plütschow A, Fuchs M. Long-Term Course of Patients With Stage IA Nodular Lymphocyte-Predominant Hodgkin Lymphoma: A Report From the German Hodgkin Study Group. J. Clin. Oncol. 2015; 33:2857–2862.
Kenderian SS, Habermann TM, Macon WR. Large B-cell transformation in nodular lymphocyte-predominant Hodgkin lymphoma: 40-year experience from a single institution. Blood. 2016; 127: 1960–1966.
Shankar AG, Roques G, Kirkwood AA. Advanced stage nodular lymphocyte predominant Hodgkin lymphoma in children and adolescents: Clinical characteristics and treatment outcome—A report from the SFCE & CCLG groups. Br. J. Haematol. 2017; 177:106–115.
Marks LJ, Pei Q, Bush R. Outcomes in intermediate-risk pediatric lymphocyte-predominant Hodgkin lymphoma: A report from the Children’s Oncology Group. Pediatr. Blood Cancer. 2018: 65:e27375.
Binkley MS, Rauf MS, Milgrom SA. Stage I-II nodular lymphocyte-predominant Hodgkin lymphoma: A multi-institutional study of adult patients by ILROG. Blood. 2020; 135:2365–2374.
Eichenauer DA, Plütschow A, Fuchs M. Long-Term Follow-Up of Patients with Nodular Lymphocyte-Predominant Hodgkin Lymphoma Treated in the HD7 to HD15 Trials: A Report From the German Hodgkin Study Group. J. Clin. Oncol. 2020; 38:698–705.
PascholdL, WillscherE, BeinJ. Evolutionary clonal trajectories in nodular lymphocyte-predominant Hodgkin lymphoma with high risk of transformation. Haematologica. 2021; 106(10):2654-2666.
Fan Z, Natkunam Y, Bair E. Characterization of variant patterns of nodular lymphocyte predominant Hodgkin lymphoma with immunohistologic and clinical correlation. Am J Surg Pathol. 2003; 27(10):1346–1356
Hartmann S, Eichenauer DA, Plutschow A. The prognostic impact of variant histology in nodular lymphocyte-predominant Hodgkin lymphoma: a report from the German Hodgkin Study Group (GHSG). Blood. 2013; 19(122): 4246-4252
Hartmann S, EichenauerDA. Nodular lymphocyte predominant Hodgkin lymphoma: pathology, clinical course and relation to T-cell/histiocyte rich large B-cell lymphoma. Pathology. 2020; 52(1):142-153. 2013
Xia D, Sayed S, Moloo Z. Geographic variability of nodular lymphocyte-predominant Hodgkin Lymphoma. Am J Clin Pathol. 2022; 157:231–43.
El Weshi A, Akhtar S, Mourad W. T-cell/histiocyte-rich B-cell lymphoma: Clinical presentation, management and prognostic factors: Report on 61 patients and review of literature. Leuk. Lymphoma. 2007; 48: 1764–1773.
Bouabdallah, R, Mounier N, Guettier C. T-Cell/Histiocyte-Rich Large B-Cell Lymphomas and Classical Diffuse Large B-Cell Lymphomas Have Similar Outcome After Chemotherapy: A Matched-Control Analysis. J. Clin. Oncol. 2003; 21: 1271–1277
Franke S, Wlodarska I, Maes B. Lymphocyte predominance Hodgkin disease is characterized by recurrent genomic imbalances. Blood. 2001; 97(6):1845–1853
Franke S, Wlodarska I, Maes B. Comparative genomic hybridization pattern distinguishes T-cell/histiocyte-rich B-cell lymphoma from nodular lymphocyte predominance Hodgkin's lymphoma. American Journal of Pathology. 2002; 161: 1861– 1867.
Goltsev Y, Samusik N, Kennedy-Darling J. Deep Profiling of Mouse Splenic Architecture With CODEX Multiplexed Imaging. Cell. 2018; 174(4): 968–81.
Schurch CM, Bhate SS, Barlow GL. Coordinated Cellular Neighborhoods Orchestrate Antitumoral Immunity at the Colorectal Cancer Invasive Front. Cell. 2020; 182(5): 1341–59.e19.
Phillips D, Matusiak M, Gutierrez BR. Immune Cell Topography Predicts Response to PD-1 Blockade in Cutaneous T Cell Lymphoma. Nature Comm. 2021; 12(1): 6721
NguyenQH, PervolarakisN, BlakeK. Highly Multiplexed Phenotyping of Immunoregulatory Proteins in the Tumor Microenvironment by CODEX Tissue Imaging. front Immunol. 2021; 19(12): 687673.
Du Z, Lin JR, Rashid R. Qualifying antibodies for image-based immune profiling and multiplexed tissue imaging. Nat Protoc; 2019:14 (10): 2900-2930
Spidlen J, Moore W, Parks D. Data File Standard for Flow Cytometry, version FCS 3.1. Cytometry A. 2010; 77(1): 97-100.
Satija R, Farrell J.A, Gennert D. “Spatial reconstruction of single-cell gene expression data.” Nature Biotechnology. 2015; 33: 495-502
Butler A, Hoffman P, Smibert P. “Integrating single-cell transcriptomic data across different conditions, technologies, and species.” Nature Biotechnol. 2018; 33-37
Stuart T, Butler A, Hoffman P. “Comprehensive Integration of Single-Cell Data.” Cell. 2019; 177: 1888-1902.
Hao Y, Hao S, Andersen-Nissen E. “Integrated analysis of multimodal single-cell data.” Cell. 2021; 184(13): 3573-3587
Brune V, Tiacci E, Pfeil I. Origin and pathogenesis of nodular lymphocyte-predominant Hodgkin lymphoma as revealed by global gene expression analysis. J Exp Med. 2008; 205(10): 2251–2268
Hartmann S., Tousseyn T., Doring C. Macrophages in T cell/histiocyte rich large B cell lymphoma strongly express metal-binding proteins and show a bi-activated phenotype. Int J Cancer. 2013; 133: 2609-2618. 2013B
Zhang X, Tsang P. IMP3/KOC1, a new immunohistochemical marker for differentiating classical Hodgkin Lymphoma and nodular lymphocyte predominant Hodgkin lymphoma from diffuse large B-Cell lymphoma. Am J Clin Pathol. 2018; 149:83-S4.
Kosari F, Bakhshi T, Ameli F. The utility of IMP3 immunohistochemical staining in differentiating nodular lymphocyte predominant Hodgkin Lymphoma from T-Cell/Histiocyte-Rich large B-Cell lymphoma. BMC Cancer. 2022; 22: 1359
Poppe B, De Paepe P, Michaux L. PAX5/IGH rearrangement is a recurrent finding in a subset of aggressive B-NHL with complex chromosomal rearrangements. Genes Chromosomes Cancer. 2005; 44: 218-223
HartmannS, Döring C, Vucic E. Array comparative genomic hybridization reveals similarities between nodular lymphocyte predominant Hodgkin lymphoma and T cell/histiocyte rich large B cell lymphoma. Br J Haematol. 2015; 169(3): 415-22
Hartmann S, Schuhmacher B, Rausch T, Fuller L, Döring C, Weniger M. Highly recurrent mutations of SGK1, DUSP2 and JUNB in nodular lymphocyte predominant Hodgkin lymphoma. Leukemia. 2016; 30: 844– 53.
Song JY, Egan C, Gong Q. Mutation Spectrum of Transformed Nodular Lymphocyte Predominant Hodgkin Lymphoma and Comparison with De Novo Diffuse Large B-Cell Lymphoma. Blood. 2017; 130 (Supplement 1): 1463.
Schuhmacher B., Bein J., Rausch T. JUNB, DUSP2, SGK1, SOCS1 and CREBBP are frequently mutated in T-cell/histiocyte-rich large B-cell lymphoma. Haematologica. 2019; 104: 330-337
Contu F, Rangel-Pozzo A, Trokajlo P. Distinct 3D structural patterns of Lamin A/C expression in Hodgkin and reed-Sternberg cells. Cancers (Basel). 2018; 10(9): 286.
Bein J, Thurner L, Hansmann ML. Lymphocyte predominant cells of nodular lymphocyte predominant Hodgkin lymphoma interact with rosetting T cells in an immunological synapse. Am J Hematol. 2020; 95: 1495– 502.
Veldman J, Alsada ZND, van den Berg A. Soluble PD-L1 is a promising disease biomarker but does not reflect tissue expression in classic Hodgkin lymphoma. Br J Haematol. 2021; 193: 506– 14.
Nam-Cha S.H., Roncador G., Sanchez-Verde L. PD-1, a follicular T-cell marker useful for recognizing nodular lymphocyte-predominant Hodgkin lymphoma. Am J Surg Pathol. 2008; 32: 1252-1257
Churchill HR, Roncador G, Warnke R.A. Programmed death 1 expression in variant immunoarchitectural patterns of nodular lymphocyte predominant Hodgkin lymphoma: comparison with CD57 and lymphomas in the differential diagnosis. Hum Pathol. 2010; 41: 1726– 34.
AbukhiranI, SyrbuS.I, Holman C.J. Markers of Follicular Helper T Cells Are Occasionally Expressed in T-Cell or Histiocyte-Rich Large B-Cell Lymphoma, Classic Hodgkin Lymphoma, and Atypical Paracortical Hyperplasia. A Diagnostic Pitfall For T-Cell Lymphomas of T Follicular Helper Origin. Am J Clin Pathol. 2021; 156(3):409-426.
Tousseyn T, De Wolf-Peeters C. T cell/histiocyte-rich large B-cell lymphoma: an update on its biology and classification. Virchows Arch. 2011: 459: 557– 63.
Deli AS , Scharf S, Steiner Y , . 3D analyses reveal T cells with activated nuclear features in T-cell/histiocyte-rich large B-cell lymphoma. Modern Pathology. 2022; 35:1431–1438
Vermare A, GuérinM, Peranzoni E. Dynamic CD8+ T Cell Cooperation with Macrophages and Monocytes for Successful Cancer Immunotherapy. Cancers (Basel). 2022; 14(14): 3546.
Van Goethem E., Poincloux R., Gauffre F. Matrix architecture dictates three-dimensional migration modes of human macrophages: differential involvement of proteases and podosome-like structures. J Immunol. 2010; 184: 1049-1061
Ninomiya S, Hara T, Tsurumi H . Indoleamine 2,3-dioxygenase in tumor tissue indicates prognosis in patients with diffuse large B-cell lymphoma treated with R-CHOP. Ann Hematol. 2011; 90(4):409–416
Eladl AE, Satou A, Elsayed AA. Nodular lymphocyte predominant Hodgkin lymphoma: Clincopathological study of 25 cases from Japan with a reappraisal of tissue associated macrophages. Pathol Int. 2015; 65: 652-660
Hartmann S, Soltani AS, Bankov K. Tumour cell characteristics and microenvironment composition correspond to clinical presentation in newly diagnosed nodular lymphocyte-predominant Hodgkin lymphoma. British J Hematol. 2022; 199(3):382-391
Autio M, Leivonen S, Bruck O. Clinical Impact of Immune Cells and Their Spatial Interactions in Diffuse Large B-Cell Lymphoma Microenvironment. Clin Cancer Res. 2022; 28:781–92
Chemnitz JM, Eggle D, Driesen J. RNA fingerprints provide direct evidence for the inhibitory role of TGF and PD-1 on CD4 T cells in Hodgkin lymphoma. Blood. 2007; 110(9):3226-33
Sattarzadeh A, Visser L, Rutgers B. Characterization of the Microenvironment of Nodular Lymphocyte Predominant Hodgkin Lymphoma. Int. J. Mol. Sci. 2016; 17: 2127
Rahemtullah A, Harris NL, Dorn ME. Beyond the lymphocyte predominant cell: CD4+CD8+ T-cells in nodular lymphocyte predominant Hodgkin lymphoma. Leukemia & Lymphoma . 2008; 49(10): 1870-1878
Visser L, Rutgers B, Diepstra A. Characterization of the microenvironment of nodular lymphocyte predominant Hodgkin lymphoma. Int J Mol Sci. 2016; 17(12): 2127.
Mita Y , Kimura M.Y , Hayashizaki K. Crucial role of CD69 in anti-tumor immunity through regulating the exhaustion of tumor-infiltrating T cells. International Immunology. 2018; 30(12): 559–567

Yes there is potential conflict of interest. S.Y. reports research funding from Kite Pharma and California Institute for Regenerative Medicine; Y.N. reports consultancy from Leica Biosystems and Roche Diagnostics, and research funding from Kite Pharma; AK performed the entire study while at Stanford. Author recently joined Akoya Biosciences team.

NLPHLsuppl102523.docx

Download PDF

Journal Publication

published 31 May, 2024

Read the published version in Blood Cancer Journal →

Editorial decision: revise
06 Feb, 2024
Review #3 received at journal
31 Jan, 2024
Reviewer #3 agreed at journal
31 Jan, 2024
Review #2 received at journal
10 Dec, 2023
Reviewer #2 agreed at journal
26 Nov, 2023
Reviewer #1 agreed at journal
23 Nov, 2023
Reviewers invited by journal
22 Nov, 2023
Editor assigned by journal
20 Nov, 2023
Submission checks completed at journal
20 Nov, 2023
First submitted to journal
18 Nov, 2023

You are reading this latest preprint version

Spatial Phenotyping of Nodular Lymphocyte Predominant Hodgkin Lymphoma and T-cell/Histiocyte-Rich Large B-cell Lymphoma

Status:

Journal Publication

Version 1

Abstract

Figures

Key Points

INTRODUCTION

METHODS

RESULTS

DISCUSSION

Declarations

References

Additional Declarations

Supplementary Files

Status:

Journal Publication

Version 1