Search

The potential of RNA-based liquid biopsy is increasingly being recognized in diffuse large B-cell lymphoma (DLBCL), the most common subtype of non-Hodgkin's lymphoma. This study explores the cell-free transcriptome in a humanized DLBCL patient-derived tumor xenograft (PDTX) model. Blood plasma samples (n = 171) derived from a DLBCL PDTX model, including 27 humanized (HIS) PDTX, 8 HIS non-PDTX, and 21 non-HIS PDTX non-obese diabetic (NOD)-scid IL2Rgnull (NSG) mice were collected during humanization, xenografting, treatment, and sacrifice. The mice were treated with either rituximab, cyclophosphamide, doxorubicin, vincristine, and prednisone (R-CHOP), CD20-targeted human IFNα2-based AcTaferon combined with CHOP (huCD20-Fc-AFN-CHOP), or phosphate-buffered saline (PBS). RNA was extracted using the miRNeasy serum/plasma kit and sequenced on the NovaSeq 6000 platform. RNA sequencing data of the formalin-fixed paraffin-embedded (FFPE) tissue and blood plasma samples of the original patient were included. Flow cytometry was performed on immune cells isolated from whole blood, spleen, and bone marrow. Bulk deconvolution was performed using the Tabula Sapiens v1 basis matrix. Both R-CHOP and huCD20-Fc-AFN-CHOP were able to control tumor growth in most mice. Xenograft tumor volume was strongly associated with circulating tumor RNA (ctRNA) concentration ( p < 0.001, R = 0.89), as well as with the number of detected human genes ( p < 0.001, R = 0.79). Abundance analysis identified tumor-specific biomarkers that were dynamically tracked during tumor growth or treatment. An 8-gene signature demonstrated high accuracy for assessing therapy response (AUC 0.92). The tumoral gene detectability in the ctRNA of the PDTX-derived plasma was associated with RNA abundance levels in the patient's tumor tissue and blood plasma ( p < 0.001), confirming that tumoral gene abundance contributes to the cell-free RNA (cfRNA) profile. Decomposing the transcriptome, however, revealed high inter- and intra-mouse variability, which was lower in the HIS PDTX mice, indicating an impact of human engraftment on the stability and profile of cfRNA. Immunochemotherapy resulted in B cell depletion, and tumor clearance was reflected by a decrease in the fraction of human CD45+ cells. Lastly, bulk deconvolution provided complementary biological insights into the composition of the tumor and circulating immune system. In conclusion, the blood plasma-derived transcriptome serves as a biomarker source in a preclinical PDTX model, enables the assessment of biological pathways, and enhances the understanding of cfRNA dynamics.

CD70 is an attractive target for chimeric antigen receptor (CAR) T-cell therapy for the treatment of both solid and liquid malignancies. However, the functionality of CD70-specific CAR T cells is modest. We optimized a CD70-specific VHH-based CAR (nanoCAR). We evaluated the nanoCARs in clinically relevant models in vitro, using co-cultures of CD70-specific nanoCAR T cells with malignant rhabdoid tumor organoids, and in vivo, using a diffuse large B-cell lymphoma patient-derived xenograft (PDX) model. Although the nanoCAR T cells were highly efficient in organoid co-cultures, they showed only modest efficacy in the PDX model. We determined that fratricide was not causing this loss in efficacy but rather CD70 interaction in cis with the nanoCAR-induced exhaustion. Knocking out CD70 in nanoCAR T cells using CRISPR/Cas9 resulted in dramatically enhanced functionality in the diffuse large B-cell lymphoma PDX model. Through single-cell transcriptomics, we obtained evidence that CD70 knockout CD70-specific nanoCAR T cells were protected from antigen-induced exhaustion. In addition, we demonstrated that wild-type CD70-specific nanoCAR T cells already exhibited signs of exhaustion shortly after production. Their gene signature strongly overlapped with gene signatures of exhausted CAR T cells. Conversely, the gene signature of knockout CD70-specific nanoCAR T cells overlapped with the gene signature of CAR T-cell infusion products leading to complete responses in chronic lymphatic leukemia patients. Our data show that CARs targeting endogenous T-cell antigens negatively affect CAR T-cell functionality by inducing an exhausted state, which can be overcome by knocking out the specific target., (©2024 American Association for Cancer Research.)