Optimizing Lentiviral Vectors For Allogeneic CAR-T Manufacturing

By Emmanuel Mensah, Keck Graduate Institute

Multiple myeloma is an aggressive cancer that disrupts the bone marrow’s ability to produce healthy plasma cells. Despite advances in immunotherapy, the disease remains a leading cause of morbidity and mortality among hematologic cancers.

Over the past decade, CAR T cell therapies have emerged as one of the most promising treatment options for patients with relapsed or refractory multiple myeloma. These therapies, however, are currently limited to autologous platforms, that is, treatments derived from the patient’s own cells. While autologous therapies have shown remarkable clinical responses, they are constrained by long lead times, high costs, and the reality that not every patient survives the waiting period between cell collection and infusion.

That unmet need formed the basis of our team’s project. We asked ourselves: “What if we could design an allogeneic, off-the-shelf CAR-T manufacturing process?” Such a process could transform accessibility and scalability, reducing patient wait times while lowering costs and expanding global reach. By sourcing cells from healthy donors and building a robust, closed, and automated manufacturing platform, we envisioned a process where a single donor could provide therapeutic doses for hundreds of patients.

Engineering The Viral Vector

At the core of CAR T cell therapy is the lentiviral vector (LVV). This vector carries the CAR transgene into donor T cells, enabling them to recognize and destroy cancer cells. Historically, anion exchange chromatography (AEX) has been the standard purification method for LVVs. However, AEX has limitations in purity and scalability.

A group of graduate students, including me, at the Keck Graduate Institute explored an alternative: VSV-G affinity chromatography. By targeting the VSV-G protein used to pseudotype lentivirus, this method achieved:

• &GT200-fold host cell protein removal,
• ~80% recovery, and
• gentle elution conditions that preserved viral integrity.

The downside is cost; the resin is nearly 50 times more expensive than AEX. But in the context of an allogeneic process where thousands of doses may be generated from a single batch, the higher up-front material cost is justified by downstream efficiencies and regulatory confidence.

Constructing The CAR-T Manufacturing Process

Designing an allogeneic CAR-T manufacturing process required us to carefully consider every stage, from donor cell collection to cryopreserved drug product. Each step had to balance yield, product quality, regulatory compliance, and scalability.

To optimize both cost and operational efficiency, we made the strategic decision to outsource the lentiviral production process and the donor apheresis step. These stages are highly specialized, capital-intensive, and already supported by contract manufacturing organizations (CMOs) with validated capabilities. Outsourcing allowed us to reduce capital investment, streamline our facility design, and focus internal resources on the steps that most directly define the value of the allogeneic.

From there, the workflow unfolded as follows:

• Isolation and enrichment – Using Cytiva’s Sefia Select system, T cells are purified with magnetic bead-based affinity selection in a closed, automated format.
• CRISPR gene editing – This is a critical step in generating universal donor cells. With MaxCyte’s GTx electroporation platform, CRISPR-Cas9 is used to disrupt two key genes:
  • TRAC locus to eliminate the native T cell receptor, thereby reducing the risk of graft-versus-host disease
  • β2M gene to prevent expression of MHC Class I molecules, reducing recognition and rejection by the host immune system.
• Transduction is accomplished through spinoculation to integrate the CAR construct via lentiviral vectors.
• Expansion takes place in Cytiva’s Sefia Expansion system, reaching ~21-fold amplification in a GMP-compliant closed environment.
• Harvest and washing – Expanded cells are harvested and washed to remove residual media, cytokines, and viral components.
• Quality inspection – Prior to formulation, in-process and final product testing are performed.
• Formulation and fill/finish are performed with Parker Hannifin’s SciLog Cryobag Filler, which enabled high-throughput automated cryobag filling for downstream cryopreservation.
• Cryopreservation and storage – Controlled-rate freezing ensures long-term stability, with final storage in vapor-phase liquid nitrogen.

Embedding Quality By Design

From the outset, we made quality by design (QbD) a cornerstone of our approach. We began by defining a quality target product profile (QTPP), capturing key attributes such as potency, persistence, and safety. We then identified critical quality attributes (CQAs) including sterility, viral safety, and potency and mapped them against critical process parameters (CPPs) such as transduction efficiency, culture temperature, and expansion time.

Risk analysis was essential. We employed fishbone diagrams and failure modes and effects analysis (FMEA) to identify potential risks, such as adventitious agent contamination or variability in expansion kinetics. Mitigation strategies were designed into the process, ensuring consistency and regulatory confidence. This QbD framework transformed our process from a series of technical steps into a coherent risk-managed strategy.

Modeling With SuperPro Designer

To ensure our design was grounded, we used SuperPro Designer to conduct mass and energy balance modeling. This enabled us to quantify raw material and media usage across each unit operation, calculate cell yields at each stage of expansion and purification, identify bottlenecks in equipment utilization, and model batch scheduling to support continuous year-round production.

Economics And Validation

No manufacturing plan is complete without financial modeling. We evaluated both CAPEX and OPEX to assess long-term viability. On the compliance side, we developed a validation master plan that mapped out IQ, OQ, and PQ for all major unit operations. This ensured alignment with ICH Q7 and 21 CFR 211, while embedding change control procedures to manage future adaptations.

The Impact: Treating Thousands More

Perhaps the most compelling result of our project is its patient impact. By establishing a repeatable, scalable process, one donor’s leukapheresis could treat up to 450 patients. Extrapolated across multiple donors and batches, this translates to over 40,000 patients annually. This outcome represents more than scientific progress; it is a step toward equity. It ensures that CAR-T therapy can move beyond small patient populations and become accessible to the tens of thousands who need it each year.

Conclusion: A Blueprint For The Future

What began as a classroom project evolved into a blueprint for the future of cell therapy manufacturing. By integrating advanced bioprocess technologies, embedding QbD principles, and rigorously modeling facility and economic feasibility, we demonstrated that an allogeneic CAR-T platform is both scientifically feasible and economically sustainable. The path forward will require collaboration between academia, industry, regulators, and technology providers. But our work shows that the building blocks are already in place. Off-the-shelf CAR-T therapy is no longer a distant dream. With scalable platforms, robust economics, and a commitment to patient access, it is within reach.

Author’s note This project was the result of a team effort. I am deeply grateful to my teammates: Farnoush Sohbati, Nigel Reed, Leo Tavormina, and David Guthary. Together, as the Starbound Team, we combined diverse perspectives and skills to develop a comprehensive, innovative blueprint for allogeneic CAR-T manufacturing.

About The Author:

Emmanuel Mensah graduated in May 2025 with a Master of Engineering in biopharmaceutical processing from Keck Graduate Institute. As a member of the Starbound project team, he contributed to process development, quality by design implementation, and techno-economic modeling to design a scalable allogeneic CAR-T manufacturing platform. Emmanuel also collaborated with Pfizer on a mechanistic modeling project focused on hydrophobic interaction chromatography, where his team developed a digital twin to optimize monoclonal antibody purification. His academic and professional interests lie at the intersection of advanced bioprocess engineering, automation, and regulatory strategy, with a focus on accelerating the accessibility of cell and gene therapies. He is passionate about translating innovation into impact and envisions a future where transformative therapies are available to patients worldwide, regardless of geography or cost.