Single-Cell Profiling Reveals Distinct Tumor Subtypes and Their Associated T-Cell Environments in Follicular Lymphoma
Bibliographic record
Abstract
Abstract Follicular lymphoma (FL) is an indolent, but incurable malignancy as most patients eventually experience progressive disease. We hypothesized that clonal heterogeneity and patient-specific immune responses would contribute to variable clinical outcomes and that understanding the complexity of the entire tumor "ecosystem" would allow us to better match patients with specific types of tumor- and immune-targeted therapies. In this study, we performed 38-dimensional single-cell phenotyping by mass cytometry (CyTOF) to simultaneously characterize both the substructure of malignant B cell populations as well as the T cell microenvironment in a cohort of 77 diagnostic patient FL biopsies and 35 benign reactive LN (rLN) biopsies. We first applied the t-distributed Stochastic Neighbour Embedding (t-SNE) algorithm to explore intra- and inter- tumoral heterogeneity among malignant B cell populations. t-SNE mapping of individual samples showed that more than a third of FL samples contain at least two phenotypically distinct tumor subpopulations, supporting the notion of multi-clonal tumor architectures presumably due to ongoing clonal evolution. Batched analysis combining all 77 FL cases together with 35 rLN samples revealed two distinct tumor subtypes comprising about 25% (type "A") and 10% (type "B") of total FL samples, respectively, with individual tumors within each subtype showing highly similar and partially overlapping phenotypes. Mapping the same data using Uniform Manifold Approximation and Projection (UMAP), a dimensional reduction algorithm similar to t-SNE but preserves global structure more accurately, revealed that type A tumors localized in close proximity to normal germinal center (GC) B cells, thus fulfilling conventional expectations as to the histogenesis of FL. In contrast, type B tumors localized more closely to pre-GC B cells, implying the existence of an alternate histogenic path in FL. Importantly, we also performed single-cell RNA-Seq on a subset of FL cases which independently confirmed the type A vs type B distinction in whole transcriptomic space. We next analyzed matching T cell data using a modified Statistical Scaffold algorithm in order to place distinct subsets in context with conventionally defined normal T cell populations. Clustering analysis using multi-layer phenograph performed on T cells from all FL and rLN samples combined yielded hundreds of small, but phenotypically distinct populations that were then annotated according to the nearest conventionally defined T cell subset. These imputed designations were used as features to perform hierarchical clustering of samples which revealed 3 major clusters. Cluster1 was characterized by mostly naive T cell populations and contained the majority of rLN samples. Cluster2 was characterized by more differentiated effector T cell populations and was dominated by FL samples. Samples within Cluster2 could be further divided into Tfh, Treg and Th1-rich subgroups. Cluster3 was characterized by a diverse T cell environment including naive, memory and differentiated effector subsets and contained a mixture of rLN and FL samples. Integrative analysis correlating B- and T- cell features revealed type B FL tumors were associated with a Tfh-rich immune landscape. Taken together, these data reveal pervasive phenotypic heterogeneity in both malignant and immune cell compartments of patient FL samples and suggest that defining tumoral subtypes as well as the status of the local immune response within individual samples will support more refined diagnostic classification and highlight functional interactions most amenable to therapeutic targeting. Disclosures Gascoyne: NanoString: Patents & Royalties: Named Inventor on a patent licensed to NanoString Technologies. Scott:Celgene: Consultancy, Honoraria; Roche: Research Funding; NanoString: Patents & Royalties: Named Inventor on a patent licensed to NanoString Technologies, Research Funding; Janssen: Research Funding. Steidl:Juno Therapeutics: Consultancy; Seattle Genetics: Consultancy; Nanostring: Patents & Royalties: patent holding; Bristol-Myers Squibb: Research Funding; Tioma: Research Funding; Roche: Consultancy.
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 | 0.000 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.001 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 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".