Postdoctoral Scholar Northwestern University, United States
Introduction:: The autologous transplantation of tumor-infiltrating lymphocytes (TILs) derived from resected tumors has emerged as a promising therapeutic approach in the clinic. TIL-based adoptive cell therapy (ACT) offers distinct advantages over other allogenic and engineered cell therapies due to its inherent heterogeneity, which maximizes the tumor-recognizing T-cell receptors (TCRs) while minimizing off-tissue effects. Notably, TILs have demonstrated highly encouraging clinical outcomes, with long-term complete responses observed in subsets of melanoma patients. However, while TIL-mediated ACT has primarily shown applicability for resectable metastatic melanoma, which provides large ( >3 cm) lesions suitable for TIL isolation, the accessibility of such large lesions is limited for other solid tumors. Additionally, certain patients may not be eligible for excisional surgery due to high risk or rapid tumor progression. Therefore, it is critical to develop a new workflow to acquire TIL or TIL-like cells in a minimally invasive manner to benefit a broader group of patients.
Previous studies have confirmed the presence of circulating tumor-reactive lymphocytes (CTRLs) at very low frequencies in the circulation (as low as around 0.002%). Given the intriguing possibility of non-invasively isolating tumor-reactive cells from the blood, numerous studies have aimed to isolate these rare cells for characterization and therapeutic administration. However, none of the conventional cell sorting techniques had success in isolating such a rare cell population. Here, we developed a high-performance immunomagnetic microfluidic approach that is customized for the purification of CTRLs. We successfully isolated CTRLs for the first time and systematically evaluated their phenotypes and therapeutic potency [1-2].
Materials and Methods:: Chip making and operation: The microfluidic chips for CTRL purification were generated by a standard soft lithography process. In brief, the master mold was 3D printed by a stereolithographic 3D printer with a layer thickness of 25 μm. The actual chips were made by casting polydimethylsiloxane (PDMS) on printed molds, followed by curing, peeling off, and plasma bonding to thin coverslips. The chips were degassed by overnight pluronic treatment before use. The flow rate was optimized to 16 mL/hr for the isolation process by a preliminary run using RBC-lysed blood samples.
Flow cytometry and CyTOF: Single-cell suspension from chip sorting was submitted to the centre for advanced single-cell analysis at the SickKids Research Institute for flow cytometry and CyTOF. For flow cytometry, acquired data were gated through FSC-SSC and single cell through FlowJo v 10.5. For CyTOF, acquired data were processed by gating center, width, residual, and 193Lr-DNA2 channel.
Expansion of CTRLs and animal models: CTRLs isolated from the blood were co-cultured with the irradiated feeder CD8+ T cells isolated from the spleen, at the donor:feeder ratio of 1:50 in IMDM with 10% FBS, 2X GlutaMAX, 500 ng/mL mouse interleukin 2, 25 ng/mL mouse interleukin 15 and 10 ng/mL recombinant mouse TGF-beta. Post expansion, 1 x 106 expanded CTRLs were introduced intravascularly through tail-vein injection to tumor-bearing mice. Tumor growth was monitored twice a week starting from day 5 for 40 days by caliper. Mice were euthanized at the endpoint for immunohistological analysis.
Results, Conclusions, and Discussions:: Results: We first used animal models with defined tumor epitopes to confirm the presence of CTRLs and optimize the isolation procedure. We found that the initial abundance of CTRLs is extremely rare (as low as 0.0002%). Therefore, we designed microfluidic devices with ‘X’-shaped capture structures to maximize the capture efficiency of rare cells (Figure 1A). Upon optimization, we achieved a purity of CTRLs up to 40% post one round of microfluidic assay (Figure 1B). By processing a large volume of blood (10 mL) simultaneously, we captured a sufficient number of CTRLs for sequencing and cytometry. Multi-omic data revealed that CTRLs strongly express CD103 on their membrane (Figure 2A) and therefore CD103 can be used as a specific biomarker for CTRL enumeration.
Subsequently, we isolated CTRLs using microfluidics based on the expression CD103, developed a feeder-based protocol to expand isolated CTRLs, and put them back into the tumor-bearing mice for therapeutic assessment. We found that CTRLs, although isolated from blood, have a strong ability to infiltrate into tumors for tumor killing and extended the median survival time by 40% in mouse B16 melanoma models (Figure 2B). CTRLs could also be used with the existing immune checkpoint blockade as a cocktail. In mouse MC38 colon cancer models, the cocktail completely eliminated the tumors in 80% of the mice ((Figure 2C) and generated long-lasting immunity against the same tumor type.
Conclusions: Taken together, we show that rare tumor-reactive cells from peripheral blood can be isolated in a non-invasive manner with a high-performance microfluidic system. We further demonstrated that the CTRLs, upon isolation and expansion, have comparable therapeutic potency to TILs and response to immune checkpoint blockades in animal models.
Discussions: The microfluidics-assisted discovery of CTRL in blood circulation highlights a new strategy for isolating therapeutic cells for ACT. The minimal invasiveness of blood collection makes CTRL isolation a more feasible and amenable process for patients compared to the invasive TIL acquisition. The implementation of CTRLs would greatly extend the applicability of ACT and may provide a new treatment option for late-stage patients with unresectable and/or metastasized tumors.
Acknowledgements (Optional): : This research was part of the University of Toronto’s Medicine by Design initiative, which receives funding from the Canada First Research Excellence Fund. The study was also supported in part by the McCormick Catalyst Fund at Northwestern University.
References (Optional): : [1] Z. Wang, S. Ahmed, M. Labib, H. Wang, X. Hu, J. Wei, Y. Yao, J. Moffat, E.H. Sargent, S.O. Kelley. Efficient recovery of potent tumour-infiltrating lymphocytes through quantitative immunomagnetic cell sorting, Nature Biomedical Engineering, 2022, 6, 108-117.
[2]Z. Wang, S. Ahmed, M. Labib, H. Wang, L. Wu, F. Bavaghar-Zaeimi, N. Shokri, S. Blanco, S. Karim, K. Czarnecka, E. Sargent, R. McGray, M. de Perrot, S. Kelley. Isolation of tumour-reactive lymphocytes from peripheral blood via microfluidic immunomagnetic cell sorting, Nature Biomedical Engineering, published online.
