Predicting Cell Of Origin In Diffuse Large B-Cell Lymphoma By High Dimensional Flow Cytometry
Bibliographic record
Abstract
Abstract Introduction Diffuse Large B-cell Lymphoma (DLBCL) is the most common histologic subtype of non-Hodgkin Lymphoma (NHL). Patients with DLBCL can be divided into prognostic subgroups according to the cell of origin (COO) of the lymphomas that are characterized as the germinal centre B-cell-like (GCB) and the activated B-cell-like (ABC) subgroup (Alizadeh et al Nature 2000:403:503-511). While Gene Expression Profile (GEP) is the “gold standard” method for defining COO in DLBCL, to perform GEP routinely on clinical cases would incur both financial and logistical pressures that would be challenging to overcome in many centers. As a result, several groups have endeavored to devise immunohistochemical (IHC) staining panels that recapitulate GEP results, but have met with variable success. In contrast with the inherently subjective nature of IHC result interpretation and the limited number of IHC-stained cells that can be assessed visually, flow cytometry is a widely available,mature and robust methodology that achieves quantitative measurement of multiple simultaneous antibody staining parameters on tens to hundreds of thousands of individual cells in a matter of minutes. In this study, we sought to devise a flow cytometric method for determining COO in DLBCL from freshly disaggregated patient lymph node biopsy specimens. Methods Six GCB cell lines (OCI-Ly1, SU-DHL-10, SU-DHL-4, SU-DHL-8, Karpas 422 and Pfeiffer) and 6 ABC cell lines (SU-DHL-9, NU-DUL-1, OCI-Ly3, HBL-1, MD903 and OCI-Ly10) were stained with eight cell surface markers (CD45, CD19, CD10, CD27, CD38, CD138, CD44 and CD34), two intracellular markers (BCL6 and IRF4), and a viability dye simultaneously using a single-tube, 11-color assay and analyzed on a 4-laser BD LSRFortessa flow cytometer. Antibodies were chosen to include known GCB (CD10, CD27, CD38 and BCL6) and ABC (CD138, CD44 and IRF4) markers based on prior studies. CD45 and CD19 were included to assist in identifying the malignant B cell population and CD34 was included as an investigational marker. FCS file data were analyzed using FlowJo software. Results Evaluation of mean fluorescence intensity (MFI) values and their corresponding standard deviations for 40,000-190,000 viable events for each sample revealed CD27 and BCL6 to be expressed at significantly higher levels in GCB cell lines, and CD44 and IRF4 to be significantly higher in ABC cell lines. CD10 and CD34 tended toward higher expression in GCB and ABC cell lines, respectively, but these differences did not reach statistical significance according to the Student’s t-test. Analysis of all 10 fluorescence parameters simultaneously (i.e. 10-dimensional data projection) requiring more sophisticated bioinformatic tools is currently in progress. According to the serial univariate COO decision algorithm outlined by Hans et al (Blood 2004;103:275-282), our flow cytometry results correctly assigned 5 out of 6 GCB and 4 out of 6 ABC cell lines (cell line COO designations as defined by GEP). Discussion In this study, we assessed the performance of a flow cytometry-based assay for determining COO in DLBCL. Our results demonstrated that most GCB and ABC DLBCL cell lines can be correctly assigned with initial 10-color flow panel; however, some markers were highly informative while others were less or not informative at all in COO discrimination. Further development will be needed to improve the antibody panel design to the point they are able to replicate GEP results faithfully. Although flow cytometry cannot be performed on archival pathology material (i.e. formalin fixed, paraffin embedded tissue), most community and academic pathology departments have either their own flow cytometry instrumentation on site or established local referral centers and thus fresh tissue can usually be allocated for multidimensional flow analysis. Our results support the usefulness of flow cytometry for COO determination in DLBCL. However, further optimization of the panel design and testing on patient biopsy samples is needed. Disclosures: No relevant conflicts of interest to declare.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.001 | 0.000 |
| Insufficient payload (model declined to judge) | 0.001 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".