Microfluidic technique boosts CAR T cell transduction at low viral doses

CAR-T cell therapy has transformed the treatment of certain blood cancers, but the manufacturing process behind these engineered immune cells remains slow, expensive, and demanding. Now, a research team at The Second Qilu Hospital of Shandong University in China has developed a microfluidic approach that dramatically compresses the core manufacturing timeline for CD19-targeting CAR-T cells into just 24 hours, while actually improving key quality metrics compared with conventional production. The study, published in the Journal of Translational Medicine, demonstrates that brief T-cell activation paired with closed-loop microfluidic recirculation can boost lentiviral transduction efficiency even when viral vector is scarce, a finding with potentially far-reaching implications for the cost and accessibility of cell therapy manufacturing.

Conventional CAR-T manufacturing involves harvesting a patient’s T cells, activating them, introducing the gene for the chimeric antigen receptor using a viral vector, expanding the cells over days to weeks in bioreactors, and then infusing the final product back into the patient. This prolonged ex vivo processing not only drives up costs, but also tends to push T cells toward a terminally differentiated, exhausted state, which can compromise their persistence and antitumor potency after infusion. Moreover, standard protocols typically rely on high multiplicity of infection (MOI), meaning large quantities of lentiviral vector per cell, to achieve adequate transduction rates. Since lentiviral vector is one of the most expensive components of the manufacturing workflow, reducing the required MOI could significantly lower production costs.

The Shandong University team, led by corresponding author Dongqi Tang and including Ying Jiang, Guidong Zhu, and Haiyan Zhang, hypothesized that the bottleneck might lie in how cells encounter the virus during the transduction step. In static culture plates, T cells and lentiviral particles interact largely by passive diffusion and random contact. By contrast, microfluidic systems can actively circulate cells through narrow channels, dramatically increasing the frequency and quality of cell-virus encounters within a controlled, closed environment.

To test this idea, the researchers activated primary human T cells with CD3/CD28 beads for only four hours, a far shorter activation window than conventional protocols, which often require one to three days. They then transduced these cells with a CD19 CAR/green fluorescent protein (GFP) reporter lentivirus at low MOI values of either 0.5 or 1.0. For each donor, the team ran side-by-side comparisons: one batch of cells was transduced in the microfluidic chip under closed-loop recirculation for 20 hours, while a matched batch underwent conventional static transduction in a plate. This donor-matched design allowed rigorous statistical comparison using paired t-tests and two-way repeated-measures ANOVA with Šídák correction, controlling for the substantial individual variation inherent in human donor cells.

The results were striking. At MOI 0.5, microfluidic processing achieved a Day 7 CAR/GFP positivity of 14.6 percent plus or minus 1.7 percent, compared with just 7.3 percent plus or minus 1.2 percent in matched static transduction. At MOI 1.0, the advantage persisted, with microfluidic transduction reaching 22.1 percent plus or minus 1.5 percent versus 13.9 percent plus or minus 1.7 percent in static conditions. In both cases the improvement was highly statistically significant, with p values below 0.001, and critically, the enhanced transduction came without any reduction in cell viability. Roughly speaking, the microfluidic approach nearly doubled the proportion of successfully engineered cells at both viral doses.

Perhaps more importantly from a product quality standpoint, when the researchers compared microfluidic rapid-manufactured CAR-T cells (MF-rmCAR-T), generated at MOI 1.0 with the full 24-hour core workflow, against conventionally manufactured CAR-T cells (cmCAR-T), the microfluidic products displayed several favorable attributes. Both products showed comparable CAR/GFP positivity, viability, viable-cell recovery, and Day 7 expansion, meaning the accelerated process did not sacrifice overall cell yield or engineering success. However, the MF-rmCAR-T products carried a significantly lower bulk vector copy number (VCN), averaging 1.2 plus or minus 0.2 copies per cell compared with 2.8 plus or minus 0.4 copies per cell for conventional products. This difference was statistically significant at p below 0.01, and it matters because lower VCN is generally considered safer, reducing the theoretical risks of insertional mutagenesis and excessive transgene expression.

The phenotype of the microfluidic products was also encouraging. Flow cytometric analysis revealed a greater proportion of central memory T cells, the long-lived, self-renewing subset associated with durable antitumor responses and better persistence after infusion, in the MF-rmCAR-T products. At the same time, these cells expressed lower levels of TIM-3 and LAG-3, two well-established markers of T-cell exhaustion. In the context of CAR-T therapy, exhaustion markers are closely watched quality attributes, as their upregulation during manufacturing often predicts poor clinical performance.

Functional testing reinforced these phenotypic advantages. In vitro, MF-rmCAR-T cells showed numerically higher bulk killing of Raji-Luc target cells, a lymphoma cell line engineered to express luciferase for bioluminescent tracking, and, after the researchers normalized the comparison to equivalent numbers of CAR/GFP-positive effector cells, the microfluidic products demonstrated greater lysis capacity. Cytokine profiling showed higher interferon-gamma release from the microfluidic cells following tumor encounter, indicating robust activation of the antitumor immune response. Notably, after co-culture with tumor cells, MF-rmCAR-T cells expressed lower levels of PD-1, another exhaustion marker, suggesting that the brief, microfluidic manufacturing process better preserves the functional fitness of the engineered cells.

To assess whether these in vitro advantages would translate to a living system, the team conducted exploratory experiments in a systemic xenograft model, in which immunodeficient NSG mice were engrafted with luciferase-expressing Raji lymphoma cells. Both MF-rmCAR-T and cmCAR-T products reduced tumor burden and prolonged survival compared with untransduced T-cell controls, confirming that the microfluidic process generates genuinely therapeutic cells. Although the two CAR-T groups did not differ significantly from each other in this model, the key finding stands: the dramatically shortened 24-hour manufacturing workflow produced cells with antitumor activity fully on par with conventional products, while requiring less viral vector and yielding a cleaner, less exhausted cellular phenotype.

The implications of this work extend beyond laboratory efficiency. A 24-hour core manufacturing process, compared with the multi-day to multi-week timelines of conventional protocols, could reduce facility costs, minimize the window for contamination, and open the door to truly point-of-care or fully closed, automated manufacturing systems. Because the microfluidic approach maintains high transduction efficiency at low MOI, it directly addresses one of the largest cost drivers in CAR-T production: the expense of lentiviral vector, which can account for a substantial fraction of the total cost of goods. Lower vector consumption would be especially transformative for decentralized and regional manufacturing models, where batch sizes are small and per-batch vector costs weigh heavily.

The technology also aligns with broader trends in cell therapy manufacturing, where closed, automated, and miniaturized systems are increasingly favored to reduce operator error, maintain sterility, and enable reproducible scaling. The PDMS-based microfluidic chips used in this study provide a controlled microenvironment in which cells are continuously recirculated through the viral suspension, ensuring uniform exposure across the cell population. This mechanistic advantage, enhanced contact frequency between T cells and viral particles without harsh centrifugation or chemical enhancement agents such as protamine sulfate or spinoculation, may explain both the improved transduction efficiency and the preserved cell health.

The authors are careful to frame their findings as early-stage. The study used healthy donor T cells rather than patient-derived cells, which often carry a more exhausted and less proliferative baseline state, and the animal model was exploratory rather than designed to power subtle comparisons between manufacturing methods. The team notes that these findings support evaluation in larger preclinical studies and in scalable closed manufacturing systems before any clinical translation. Larger numbers of donors, patient samples, and disease contexts will be needed to confirm that the microfluidic workflow is robust across the heterogeneity of real-world starting materials.

Still, the study represents a meaningful step toward faster, cheaper, and potentially higher-quality CAR-T manufacturing. By demonstrating that a 24-hour core process combining brief CD3/CD28 activation with closed-loop microfluidic recirculation can deliver improved low-MOI transduction, lower vector copy number, a more favorable central memory phenotype, reduced exhaustion markers, and preserved antitumor function, the Shandong University team has provided a compelling proof of concept. If subsequent studies validate these results at scale, microfluidic rapid manufacturing could help address one of the central paradoxes of modern cancer immunotherapy: that the most effective engineered cell therapies remain among the most difficult and expensive medicines to produce. The work was supported by the Major Scientific and Technological Innovation Project of Shandong Province, and the study was approved by the Ethics Committee of the Second Qilu Hospital of Shandong University.