Yongke Zhang, CSO of IASO Biotechnology

In vivo CAR T has been touted as the next major revolution in cell therapy. Early clinical trial results may be promising, but will they stand up to scrutiny?

Beacon had the pleasure of sitting down with Yongke Zhang, Chief Scientific Officer of IASO Biotechnology, a commercial-stage biopharma company specialised in discovery, development and manufacturing of innovative cell therapies, to understand why they believe in vivo has significant potential.

Could you briefly describe your scientific journey and what led you to become CSO of IASO Biotechnology?

I was trained as a physician at Beijing Medical University and conducted research on the use of tumor-infiltrating lymphocytes (TILs) to treat lung cancer at Chiba University in Japan for my doctor degree in clinical medicine. I then completed my PhD in Molecular and Cellular Biology at the University of Tokyo, followed by postdoctoral training at the DNAX Research Institute, now part of Merck Biologics, in the United States.

I subsequently joined Protein Design Labs, now part of AbbVie, and Epitomics in the Bay Area, where I gained extensive experience in therapeutic antibody development and antibody engineering. I later helped spin out Apexigen, where we were among the pioneers in using T-cell priming approaches for immuno-oncology.

I went on to start my own company in partnership with a Chinese pharmaceutical company, at a time when China’s biotech sector was undergoing rapid growth. Throughout my career, I have believed that cell therapy has the potential to give patients a real chance at achieving durable responses, and the early clinical results with CAR T therapy were particularly encouraging.

Those results sparked my interest because CAR T brought together two areas that had shaped my career—antibody engineering and cell therapy. This ultimately led me to IASO Biotechnology (IASO Bio), where I now focus on advancing our in vivo CAR T platform and developing the next generation of in vivo CAR T therapies.

 

At what point did you think that in vivo as opposed to ex vivo CAR T could be a transformative modality?

IASO has already developed and launched an ex vivo BCMA CAR T therapy, with remarkable clinical results. However, ex vivo approaches still face several important challenges. Patient access can be limited by the high cost of treatment, and reimbursement can be challenging in many countries. In addition, ex vivo CAR T requires a complex, patient-specific manufacturing process, rather than being an off-the-shelf therapy.

Ex vivo CAR T also typically requires lymphodepleting chemotherapy, which can be highly toxic and increase the risk of serious infections. Initially, we explored allogeneic cell sources as a potential solution. However, we found that the patient’s immune system could eliminate the allogeneic cells, requiring even more intensive lymphodepletion. We also found that CAR T cell persistence and clinical efficacy were not sufficient.

For these reasons, we decided to move away from the allogeneic approach and focus our efforts on developing an in vivo CAR T platform. The idea is to generate CAR T cells directly inside the patient, potentially overcoming many of the limitations associated with conventional ex vivo manufacturing. That realization was a turning point for us and led us to believe that in vivo CAR T could become a transformative modality for cell therapy.

 

What are the most significant scientific and technical challenges facing in vivo CAR T development?

There are several significant scientific and technical challenges. The first is delivering the CAR specifically and efficiently to T cells, which is fundamental to the success of an in vivo CAR T approach. The second is ensuring that the CAR T cells generated in vivo can persist in the body long enough to achieve a durable therapeutic effect.

A third challenge is determining whether in vivo-generated CAR T cells can reach the same functional cell numbers and activity that can be achieved with ex vivo approaches. Addressing these challenges requires careful product design, from the delivery system to the CAR construct itself, to ensure efficient T cell targeting, robust expansion, persistence, and therapeutic function.

 

Do you see a future role for both permanent and transient expression?

I think it depends on the platform and the disease setting. In oncology, published clinical data suggest that persistent CAR expression may be important for controlling tumors that continue to proliferate over time. In autoimmune disease, however, the answer is less clear.

Transient CAR expression may offer a potentially safer approach because it could avoid the risks associated with permanent genetic modification, including insertional mutagenesis. However, recent clinical data in systemic lupus erythematosus (SLE) suggest that transient approaches may have more limited efficacy compared with viral vector-based approaches.

Therefore, we need to evaluate whether transient CAR expression can provide sufficient depth and durability of response to effectively reset the immune system in patients with severe autoimmune diseases. Ultimately, I think the optimal approach may depend on both the biology of the disease and the characteristics of the underlying in vivo CAR T platform.

 

What is it about in vivo CAR T that makes autoimmune diseases particularly attractive compared to ex vivo CAR T?

We are preparing to advance our in vivo CAR T asset into clinical trials for autoimmune diseases because we believe this approach has the potential to offer a favorable safety profile compared with conventional ex vivo CAR T.

With ex vivo CAR T, a large number of engineered T cells are infused into the patient at once. This can lead to rapid and substantial expansion of CAR T cells, which may increase the risk of cytokine release syndrome (CRS) and other associated toxicities, including hemophagocytic lymphohistiocytosis (HLH), which has been reported more recently. In contrast, in vivo CAR T starts with a relatively small number of transduced T cells that expand primarily when they encounter their target antigen. This expansion may therefore be more gradual and potentially more self-limiting.

In cancer, a high antigen burden can drive substantial CAR T cell expansion, which may be important for achieving an effective antitumor response. In autoimmune disease, following B cell or plasma cell depletion, the relevant antigen burden is generally lower. As a result, CAR T cell expansion may be more controlled.

In this sense, we think of in vivo CAR T as behaving somewhat like a “smart drug”—it can expand when and where the target antigen is present, and its expansion may diminish as the antigen burden decreases. This concept is reflected in the name of our in vivo CAR T platform, InTelliCAR™.

Insertional mutagenesis is an understandable concern with integrating vectors. However, extensive clinical experience with lentiviral vector-based therapies, together with large-scale analyses, has not established a clear causal link between lentiviral vector integration and secondary malignancies. This growing body of clinical evidence has helped strengthen confidence in the safety profile of lentiviral vector approaches.

 

How would you identify a ‘winner’ in the in vivo CAR T space?

I think there are two key factors. The first is delivery specificity, because precise targeting is central to the safety of an in vivo CAR T approach. Some published studies have reported severe infusion-related reactions when the delivery system activated the innate immune system. In our own clinical trial, we did not observe that pattern. The reactions we saw were mostly limited to Grade 1 fever and resolved within 12 hours. We plan to present these data at the International Myeloma Society Annual Meeting in Glasgow and at ASH later this year.

The second factor is the ability of the CAR T cells to expand and persist in the body compared with ex vivo CAR T. In vivo expansion depends in part on the patient’s underlying T cell fitness, so there can be greater variability than with ex vivo manufacturing, where T cells can be selected, activated, and expanded before they are infused. Even at the same vector dose, the biological response can vary from patient to patient.

At the same time, we have observed important potential safety advantages with the in vivo approach, including lower rates of CRS, less cytokine-driven cytopenia, and fewer infections. These advantages may be largely attributable to the fact that in vivo CAR T does not require chemotherapy-based lymphodepletion. Ultimately, I think the success of an in vivo CAR T platform will depend on achieving the right balance of precise delivery, controlled expansion, persistence, leading to better efficacy and safety.

 

Are any concerns about in vivo CAR T overstated?

People often ask why we need an in vivo approach when ex vivo CAR T exists. It comes down to accessibility, cost, and vein-to-vein turnaround times. In vivo CAR T offers a single-dose, off-the-shelf alternative—potentially outperforming T-cell engagers on convenience.

It is still early days, and comparing early in vivo data to the long-term efficacy of ex vivo CAR T—where we see patients reaching seven years of drug-free survival—is premature. However, emerging MRD negativity and complete response rates from our IASO206 and IASO208 trials are very promising.

Unlike TCEs, which require continuous administration and risk T cell exhaustion, a single in vivo CAR T dose used earlier in treatment preserves TCEs as a fallback option upon relapse. Our goal is to move CAR T into earlier treatment lines to provide more patients with deep, durable remissions.

 

Looking ahead to 2035, how do you expect the cell therapy landscape to evolve?

Ex vivo CAR T initially entered the clinic in very late-line settings, but over time, it has moved from fourth-line to earlier lines, including frontline therapy. I expect in vivo CAR T to follow a similar trajectory because it has the potential to be easier to deliver and much safer. A one-time treatment could also improve patient convenience and compliance compared with continuous therapies like T cell engagers, where repeated dosing is required and each administration carries a recurring risk of cytokine release syndrome. If in vivo CAR T can combine strong efficacy with a simpler treatment model, it could reduce both the clinical and economic burden on patients while enabling cellular therapies to move into earlier lines of care.

 

“If in vivo CAR T can combine strong efficacy with a simpler treatment model, it could reduce both the clinical and economic burden on patients while enabling cellular therapies to move into earlier lines of care.”

 

Will in vivo CAR T displace ex vivo development?

I think there will be a place for both for some time. In vivo CAR T still needs to mature clinically and commercially—including demonstrating value in Phase 3 trials, particularly in autoimmune diseases, and against standard of care (SOC) in oncology. There is still a long way to go, but as safety and delivery efficiency improve, in vivo approaches will likely take on a broader role alongside established ex vivo therapies.

 

How important are high-quality data sources and competitive intelligence when making strategic decisions?

High-quality competitive intelligence is critical. As a research team, we need to constantly monitor the field, track competitor progress, and compare emerging data across different targets and platforms.

Platforms like Beacon are particularly valuable because they consolidate preclinical data, conference presentations, posters, and related intelligence into a single place. That unified view makes it much easier for us to monitor industry developments, select viable targets, benchmark our own programs against competitors, and identify potential partnering opportunities.

 

Where do you see the greatest opportunity for AI to accelerate innovation in radiopharmaceuticals?

AI is a powerful, practical tool for accelerating better decision-making across the drug development continuum. That is why we established an internal AI team at IASO.

Operationally, AI can streamline clinical development by supporting protocol design, site selection, EDC documentation, and administrative workflows. On the discovery side, it significantly shortens timelines—particularly in areas like de novo antibody generation and optimizing CAR architecture (including domain orientation, linker design, and hinge spacing for dual CAR constructs).

While AI helps us navigate difficult targets where traditional immunization falls short, it complements rather than replaces hands-on experimental validation and human scientific judgment. Used thoughtfully, AI allows us to move much faster while keeping human expertise at the center of innovation.

 

Finally, what is next for IASO Biotechnology?

We hit a major milestone with the US FDA clearing our IND for IASO208, our lead in vivo CAR T program, on September 16, 2026. We plan to submit an IND for our second candidate, IASO206, by year-end, followed by a third submission in Q2 of next year.

Beyond our core mono-CAR assets targeting CD19, CD20, and BCMA, our pipeline encompasses dual-CAR candidates for hematologic malignancies and autoimmune conditions. Long term, we are setting our sights on solid tumors. We hypothesize that viral vector-mediated in vivo CAR T generation can enhance T cell penetration into the solid tumor microenvironment, offering a potentially transformative alternative to traditional immunotherapies.