3D CLL spheroid model recapitulates lymph node phenotypes and in vivo responses to ibrutinib; a single-cell characterization
Ibrutinib is used to successfully treat chronic lymphocytic leukemia (CLL) patients, however patients may experience relapse during treatment, suggesting that residual tumor cells are protected at lymph node (LN) sites and subsequently may drive disease progression. An unanswered question is what happens in the LN-residing CLL cells upon ibrutinib treatment, which is not easily studied in vivo due to their scarcity after treatment-induced redistribution of CLL cells into the peripheral blood (PB). An arguably bigger drawback is that CLL is widely studied using PB samples and that many signals that drive CLL activation, proliferation and drug resistance are not recapitulated in standard in vitro models. We recently developed the first in vitro culture model for CLL that incorporates CLL, T cells and myeloid cells in a spatially organized, LN-mimicking context. This model results in a representation of CLL phenotypes that allows for better investigation of proliferation and drug resistance. In the current study, we aim to bridge the gap between in vitro and in vivo studies by investigating to what extent the 3D model mimics the in vivo LN and response to ibrutinib.
We performed single-cell RNA sequencing (scRNA-seq) on patient-derived CLL spheroids and paired peripheral blood (PB) and LN samples. First, we directly compare the 3D CLL model and corresponding ibrutinib response to matched patient samples and their in vivo response to ibrutinib. Second, we utilize the 3D model to investigate activated and proliferating CLL cells and their response to ibrutinib. Finally, we investigate the divergent ibrutinib responses observed in the 3D model between PB and LN counterparts and characterize T cells and their functional implications.
Comparison of the in vitro ibrutinib response in spheroids to the ibrutinib response in patients demonstrated that key molecular features were are also reflected in vitro. CLL spheroids amplified the proportion of the clinically relevant LN niche. Furthermore, the three known CLL fractions – resting, activated, proliferating – showed distinct ibrutinib responses. Ibrutinib treatment shifted responding CLL cells towards a resting phenotype, yet activated CLL cells showed upregulation of TLR9-MALT1-MYC signaling, suggesting emergent resistance mechanisms. Lastly, 11 T cell subsets were identified, collectively demonstrating a T cell-dependent role in promoting CLL proliferation even in the presence of ibrutinib.
Our single-cell data reveals dynamic transitions between resting, activated, and proliferating CLL phenotypes and identifies a persistent, treatment-insensitive niche with unique transcriptional adaptations which may be leveraged to guide future therapeutic interventions overcoming microenvironment-mediated resistance.