Lymphoma, Follicular

The microarray datasets and corresponding flow cytometry results were obtained from GEO (accession nos. GSE65135 and GSE65133), which included 14 disaggregated lymph node biopsie samples from patients with follicular lymphoma and 20 peripheral blood samples from individuals vaccinated for influenza, respectively.
Besides, RNA‐Seq expression profile of samples from melanoma patients and their corresponding flow cytometry of four immune cell types (B, CD4⁺ T, CD8⁺ T, and NK) result were collected from EPIC publication.[qv: 13] Because of the scarcity of bulk RNA‐Seq datasets containing both gene expression profiles and flow cytometry counts for different immune cell types, particularly for T‐cell subsets, we simulated two bulk RNA‐Seq datasets by integrating the expression profiles of seven cell types (CD4⁺ naïve, CD8⁺ naïve, MAIT, Tcm, Tex, Treg, Th) from single‐cell RNA sequencing data. The transcripts per million (TPM) normalized expression of liver and lung cancers from two Nature papers were collected from GEO (GSE98638[qv: 18] and GSE99254[qv: 19]). Based on the work of Max et al.,[qv: 43] single‐cell expression was normalized as follows:
$exp=lo{g}_2\left(exp+1\right)$
for each single‐cell dataset, the TPM values were transformed to
To ensure cross‐sample comparability, the expression of all single‐cell samples from the same dataset were normalized to the average expression of 3686 housekeeping genes[qv: 44] as follows:
$ex{p^{\prime }}_i\text{=}ex{p}_i\text{*}\frac{\overline{HK}}{H{K}_i}$
where exp_i represents the gene expression profile of sample i, HK_i denotes the average gene expression of all housekeeping genes in sample i, and HK¯ is the average expression of all housekeeping genes in all samples. Besides, a single‐cell sequencing dataset from 19 patients with melanoma was collected from GEO (accession GSE72056), which is the normalized expression matrix as described above by Tirosh et al.[qv: 20] and contains the single‐cell RNA‐Seq of B cells, T cells, macrophages, NK cells, and three other nonimmune cell types. Because CD8⁺ and CD4⁺ T cells can be easily distinguished by CD4, CD8A, and CD8B expression, we divided T cells into CD8⁺ T cells, CD4⁺ T cells, and others. Then, the bulk expression of each sample was identified by aggregating normalized expression from all cell barcodes for each patient sample. The cell ratio per cell type in a sample was calculated by the cell number of a specific cell type divided by the total number of cells (Tables S5–S7, Supporting Information).

Flow cytometric results from patients who were diagnosed with mature B-cell neoplasms from October 2015 to October 2020, were reviewed retrospectively. Each case represented a primary diagnosis of lymphoma that was made based on an incisional or excisional tissue biopsy or fine-needle aspiration biopsy specimens. Histologic slides including immunohistochemical slides, were reviewed without knowledge of the flow cytometric results to confirm the initial diagnoses in all available cases. The diagnosis was made according to the World Health Organization (WHO) 2008 classification (12 (link)), WHO 2017 classification,and WHO 2022 classification (2 , 3 (link), 13 (link)). These patients included 119 patients with DLBCL, 25 patients with Burkitt lymphoma, 67 patients with MCL, 76 patients with follicular lymphoma (FL), 30 patients with marginal zone lymphoma (MZL), 32 patients with lymphoplasmacytic lymphoma (LPL)/Waldenstrom’s macroglobulinemia (WM), 159 patients with chronic lymphocytic leukemia (CLL)/small lymphocytic lymphoma (SLL), 5 patients with hairy cell leukemia, 4 patients with mucosa-associated lymphoid tissue lymphoma (MALT-L), and 42 patients with transformed lymphoma. For the diagnosis, Ki67 expression in lymphoma cells was detected in the bone marrow, pleural effusion, and ascites or lymph node samples. The present study was approved by the Ethical Committee of Tongji Hospital, Tongji Medical College, and Huazhong University of Science and Technology (permit number TJ-IRB20200716), and all procedures conducted followed the protocols of the Declaration of Helsinki.

For diagnosing CLL and other B-cell lymphomas we used a B-cell panel which consisted of two tubes with different fluorescence antibody panels. The first tube (T1) included fluorescence antibodies against B-cell antigens (CD19, CD20, FMC7, CD79b, CD23, light chains of kappa and lambda), T-cell antigens (CD3, CD5, CD2, CD7, CD4, CD8), and the activation marker CD38, which has been described to be prognostic for CLL [14 (link)]. The second tube (T2) contained B-cell antigens (CD19, CD20, IgM), markers of hairy-cell leukemia (CD103, CD11c, CD25), follicular lymphoma, and high-grade lymphoma (CD10), and additional markers to ensure the diagnosis of CLL (CD43, CD200). The complete antibody panel, clones, and fluorescence dyes are stated in the Supplementary Information (Table S1).
Two 5 mL polystyrene FACS tubes with fluorescence antibodies in a dried-down layer (DuraClone-Technology, Beckman Coulter, Krefeld, Germany) were incubated for 15 min at room temperature in 100 μL prewashed peripheral blood. After antibody staining, red cells were lysed in 2 mL of VersaLyse™ (Beckman Coulter, Krefeld, Germany) for 10 min, washed with 3 mL of buffered phosphate saline (PBS Biochrom, Berlin, Germany), and centrifuged with 300× g for 5 min. The cell pellet was resuspended in 500 μL PBS and measured on a Navios Flow Cytometer (Beckman Coulter, Krefeld, Germany). In total, up to 1 × 10⁵ cells were acquired.