There have been two mainstream directions in the industry regarding the technological path selection of CAR-T in vivo for a long time: in vivo delivery based on lentiviral vectors and non viral delivery based on LNP (lipid nanoparticles).

In early discussions, the two were often placed in a binary opposition. As research and development deepens, more and more professionals realize that this binary oppositional thinking may deviate from the essence of technological development.

Thanks to the rich experience accumulated in the in vitro CAR-T industry over the past decade, lentiviral vectors have natural advantages in terms of cell infection efficiency and gene integration stability. Many teams that have been deeply involved in the field of in vitro CAR-T for many years have shown a fast starting speed when transitioning to in vivo applications.

On March 6th, at the "2026 In vivo CAR-T Delivery Breakthrough and Clinical Implementation Practical Seminar" jointly organized by Tongxiyi and Yinuo Biology, a biotechnology company specializing in in in in vivo modification of lentiviruses shared that its scientific founding team had started basic research on virus targeting modification and safety design as core research directions as early as 2017. This long-term accumulation enables it to quickly come up with differentiated technical solutions when the track erupts in 2024, avoiding the suspicion of simply following the trend.

However, the challenges of in vivo application of lentiviruses cannot be ignored. Directly injecting viruses into the human body requires overcoming a series of complex issues such as immune system clearance, non targeted distribution, and genome integration safety. A scientist who has been engaged in virus vector research and development for many years admitted that although the team has invested nearly four years in slow virus modification, they have still taken many detours in the process. At present, there are only a few high-quality clinical data publicly disclosed worldwide, which precisely confirms the high threshold of this technology.

The LNP route is highly sought after due to its successful application in mRNA vaccines. However, multiple experts have pointed out that there is a fundamental difference between using LNP to target T cells and targeting the liver.

LNP needs to achieve cell selective delivery while avoiding excessive immune reactions, which poses new requirements for formulation design and formulation processes. A representative of a company engaged in TCR-T research and development recalled that they had attempted to deliver TCR genes using non viral vectors more than a decade ago, but put it on hold due to the inability to achieve mass production. It was not until after COVID-19, LNP technology developed significantly, that they restarted relevant pipelines. But how to ensure the safety of repeated administration? How to maintain the persistence of gene expression? These questions still need to be answered.

It is worth noting that some investment institutions hold an open attitude towards the choice of technology routes. An investor active in the CGT field stated that simply opposing lentivirus and LNP may overlook the possibility of technological integration. Perhaps in a few years, we will discover that cleverly designed slow viruses have far superior safety performance compared to LNP. Don't be limited by fixed thinking. ”

In his view, the current stage is more important for the originality of underlying technology and the reliability of data, rather than labels in the form of carriers.

The investment logic of the CAR-T track in the body has undergone significant changes in the past two years.

In the early stages, due to the lack of clinical data, investors mainly make "left side investments" based on team background and technical concepts, that is, at the earliest stage of the technical curve and the most ambiguous stage of cognition, they bet on potential directions with relatively low valuations. This strategy has obvious risk investment characteristics and tests the judgment of trends.

As more and more companies enter the IIT (Investigator Initiated Clinical Trials) phase, the balance of investment decisions begins to tilt towards data. A senior investor described the current stage as "opening a blind box": everyone knows there are "hidden funds" in this series, but no one can predict which company will produce good results. For investors, they are both afraid of missing out on opportunities and afraid of buying at high levels at this moment.

The interpretation of IIT data itself is also full of grayscale. The head of an institution shared a recent decision-making experience: they had evaluated a project with over ten IIT data cases, but ultimately chose to abandon it. The reason is that a small sample size makes it difficult to exclude confounding factors such as baseline differences and past treatment history, and relying solely on these data is not sufficient to support group level re injection.

For companies with only one or two stunning data cases, they are more inclined to make small-scale layouts through early-stage funds rather than directly investing in listed companies.

This cautious attitude has been confirmed by clinical experts. A clinical research manager at a hospital pointed out that hospitals are increasingly inclined to require companies to "comply with IND standards" when undertaking IIT projects. Although it is a trial initiated by researchers, there are still risks involved. If the data cannot be used for subsequent registration, the research value of the hospital will be compromised. She suggests that companies communicate with clinical PIs as early as possible before initiating IIT to ensure the rigor of the trial design.

Nevertheless, the power of data cannot be underestimated. The founder of a CAR-T company that achieved positive results in IIT shared that when early data was presented at the shareholders' meeting, almost all participating parties did not hesitate to increase their investments. This indicates that in the early stages of uncertainty, real-world human data remains the most powerful persuasive tool.

As the track advances, the industry's understanding of the difficulty of clinical development is also deepening. Representatives from pharmaceutical companies with commercial product experience pointed out that the biggest challenge in the development of CAR-T in vivo lies in the design and implementation of clinical pathways.

An important variable is the dynamic evolution of standard therapy. Taking multiple myeloma as an example, a few years ago when in vitro CAR-T was approved, the standard therapy pattern was relatively simple; Nowadays, bispecific antibodies have shown strong therapeutic effects in both first-line and second-line treatments.

This means that later researchers who conduct in vivo CAR-T, regardless of the carrier used, must face head to head competition with these already marketed drugs. At present, the single arm IIT data that most companies are proud of may only be the first step in a long journey in the face of serious registration clinical trials.

The durability of therapeutic effects is another core issue.

The reason why in vitro CAR-T is known as a "curative therapy" is that it can form memory T cells and achieve deep remission for several years. This long-term therapeutic effect is currently difficult to achieve in the form of dual antibody drugs.

Therefore, whether CAR-T in vivo can reproduce the persistent characteristics of CAR-T in vitro directly determines its high clinical value.

The most realistic topic in the discussion always revolves around the payment system.

A founder of a company that has been engaged in gene therapy for many years frankly stated that although its products for rare diseases have been approved, commercialization is still struggling. In China, there is an unwritten "300000 yuan red line" for medical insurance payments - annual treatment costs exceeding 300000 yuan make medical insurance negotiations almost hopeless. The disposable nature of gene therapy naturally contradicts this payment model.

Huimin Insurance has become a breakthrough for some enterprises. But connecting the Huimin Insurance requires negotiating with local governments one by one to prove that the product will not have a significant impact on the fund pool. This process is long and full of uncertainty, and many patients are also observing: if Huiminbao can reimburse hundreds of thousands, why do they have to bear all the expenses? This kind of game may lead to the awkward situation of "being well received but not well received" in the early stages of product launch.

Compared to the US market, the pricing of rare disease gene drugs reaching millions of dollars can still be accepted by the commercial insurance system, mainly due to the decentralization of payment parties and risk sharing mechanisms. Going overseas to make money is the fundamental way, "said an investor bluntly." China's commercial insurance and medical insurance system is difficult to support expensive innovative drugs in the short term. Only by integrating into the global market can we maximize commercial value.

Some companies have already explored in this area. The founder of a pharmaceutical company deeply involved in the CGT field shared their internationalization path: they spent nearly three years breaking through policy barriers such as cross-border drug purchasing, customs supervision, and multi departmental coordination, ultimately achieving a breakthrough in Chinese production and Chinese data support for overseas approval. She suggested that colleagues should not only focus on the "volume" of domestic payment platforms, but also look at the unmet clinical needs worldwide.

Of course, internationalization does not mean giving up the Chinese market. In China, enterprises must accept the reality of "rolling", but the rolling should not be a bottomless low price, but through technological innovation and process optimization, to create true cost-effectiveness.

The issue of cost is currently the focus of the game between supply and demand. Faced with high trial and error costs, more and more companies are seeking cross-border cooperation.

A scientist engaged in T cell functional modification proposed an interesting idea: "We don't have delivery technology, but we're not in a hurry. We can wait for a delivery company to come out and then cooperate with them. We're good at solving the problem of tumor microenvironment, strengthening cell function, and combining the two to create a new product.

In a crowded race track, no company can monopolize all aspects. CDMO、CRO、 The "new infrastructure" system composed of research institutes is becoming a key support for industrial development.

The development of CAR-T in vivo is a protracted battle, and the first generation products may not be able to solve the ultimate problem, requiring iterations of the second and third generations. In this process, quality is more important than speed, and innovation in underlying logic is more important than financing.

When the blind box is opened and the data reveals its true essence, only those companies that truly respect science, clinical practice, and have a global perspective can stand out in this saturated innovation and become the ultimate survivors. For the entire industry, regardless of which company succeeds, it will be an important step towards broader applications of cell gene therapy.