Spatial transcriptomics reveals profound subclonal heterogeneity and T-cell dysfunction in extramedullary myeloma.

Extramedullary disease (EMD) is a high-risk feature of multiple myeloma (MM) and remains a poor prognostic factor even in the era of novel immunotherapies. Here we applied spatial transcriptomics (tomo-seq [n=2] and 10X Visium [n=12]), and single-cell RNA sequencing (scRNAseq [n=3]) to a set of 14 EMD biopsies to dissect the three-dimensional architecture of tumor cells and their microenvironment. Overall, the infiltrating immune and stromal cells showed both intra- and inter-patient variation with no uniform distribution over the lesion. We observed substantial heterogeneity at the copy number level within plasma cells, including the emergence of new subclones in circumscribed areas of the tumor, consistent with genomic instability. We further identified spatial expression differences of GPRC5D and TNFRSF17, two important antigens for bispecific antibody therapy. EMD masses were infiltrated by various immune cells, including T-cells. Notably, exhausted TIM3+/PD-1+ T-cells diffusely co-localized with MM cells, whereas functional and activated CD8+ T-cells showed a focal infiltration pattern along with M1 macrophages in otherwise tumor-free regions. This segregation of fit and exhausted T-cells was resolved in the case of response to T-cell engaging bispecific antibodies. MM cells and microenvironment cells were embedded in a complex network that influenced immune activation and angiogenesis, and oxidative phosphorylation represented the major metabolic program within EMD lesions. In summary, spatial transcriptomics has revealed a multicellular ecosystem in EMD with checkpoint inhibition and dual targeting as potential new therapeutic avenues.

Sequential antigen loss and branching evolution in lymphoma after CD19- and CD20-targeted T-cell-redirecting therapy.

ABSTRACT: CD19 chimeric antigen receptor (CAR) T cells and CD20 targeting T-cell-engaging bispecific antibodies (bispecs) have been approved in B-cell non-Hodgkin lymphoma lately, heralding a new clinical setting in which patients are treated with both approaches, sequentially. The aim of our study was to investigate the selective pressure of CD19- and CD20-directed therapy on the clonal architecture in lymphoma. Using a broad analytical pipeline on 28 longitudinally collected specimen from 7 patients, we identified truncating mutations in the gene encoding CD20 conferring antigen loss in 80% of patients relapsing from CD20 bispecs. Pronounced T-cell exhaustion was identified in cases with progressive disease and retained CD20 expression. We also confirmed CD19 loss after CAR T-cell therapy and reported the case of sequential CD19 and CD20 loss. We observed branching evolution with re-emergence of CD20+ subclones at later time points and spatial heterogeneity for CD20 expression in response to targeted therapy. Our results highlight immunotherapy as not only an evolutionary bottleneck selecting for antigen loss variants but also complex evolutionary pathways underlying disease progression from these novel therapies.

NetRank Recovers Known Cancer Hallmark Genes as Universal Biomarker Signature for Cancer Outcome Prediction.

Gene expression can serve as a powerful predictor for disease progression and other phenotypes. Consequently, microarrays, which capture gene expression genome-wide, have been used widely over the past two decades to derive biomarker signatures for tasks such as cancer grading, prognosticating the formation of metastases, survival, and others. Each of these signatures was selected and optimized for a very specific phenotype, tissue type, and experimental set-up. While all of these differences may naturally contribute to very heterogeneous and different biomarker signatures, all cancers share characteristics regardless of particular cell types or tissue as summarized in the hallmarks of cancer. These commonalities could give rise to biomarker signatures, which perform well across different phenotypes, cell and tissue types. Here, we explore this possibility by employing a network-based approach for pan-cancer biomarker discovery. We implement a random surfer model, which integrates interaction, expression, and phenotypic information to rank genes by their suitability for outcome prediction. To evaluate our approach, we assembled 105 high-quality microarray datasets sampled from around 13,000 patients and covering 13 cancer types. We applied our approach (NetRank) to each dataset and aggregated individual signatures into one compact signature of 50 genes. This signature stands out for two reasons. First, in contrast to other signatures of the 105 datasets, it is performant across nearly all cancer types and phenotypes. Second, It is interpretable, as the majority of genes are linked to the hallmarks of cancer in general and proliferation specifically. Many of the identified genes are cancer drivers with a known mutation burden linked to cancer. Overall, our work demonstrates the power of network-based approaches to compose robust, compact, and universal biomarker signatures for cancer outcome prediction.