Pancreatic cancer is a rare but devastating disease. Despite only accounting for around 3% of cancer cases in the US, it is responsible for 8% of all cancer deaths, with fewer than a fifth of patients expected to survive five years after their diagnosis.
One of the reasons pancreatic cancer has such a high mortality rate is that it is often discovered late. Over half of patients are diagnosed after the cancer has already metastasized, and chances of survival are at their lowest.
Scientists have speculated that symptom onset in pancreatic cancer occurs at such late stages due to pancreatic tumors’ ability to evade the immune system. This is because a layer of stromal tissue, known as the stromal barrier, promotes immunosuppression and drug resistance.
While current ex vivo chimeric antigen receptor (CAR) T-cell therapies have demonstrated tremendous success in treating hematological malignancies, ex vivo medicines feature significant drawbacks when it comes to manufacturing costs and toxicity.
This is where research from the University of Pennsylvania may be able to offer a solution. An article published in Cancer Immunology Research details an in vivo CAR T-cell approach that can target the stromal barrier utilizing mRNA encased in lipid nanoparticles (LNPs). The team found that coating the LNPs in antibodies led to specific targeting and potent anti-tumor effects that bypassed immune evasion by pancreatic cancer, all without any need for external cell manipulation.
Dismantling defenses
Corresponding authors Ellen Pure and Khuloud Bajbouj argue that the complex extracellular matrix surrounding pancreatic tumors, known as the stromal barrier, prevents immune evasion and is key to eliminating the cancer.
“Stromal barriers physically block immune cells and drugs from reaching desmoplastic solid tumors like pancreatic cancer and suppress immune responses,” Pure and Bajbouj told BioXconomy. “The stroma directly promotes tumor cell survival, growth, and metastasis and inhibits anti-tumor immunity.”
Fibroblast activation protein (FAP) is a transmembrane protease whose expression is associated with worse clinical outcomes in cancer. It is also a key component of the stromal barrier. This makes it an appealing therapeutic target in pancreatic tumors.
“We encoded an mRNA for a CAR against FAP, a marker highly expressed selectively on cancer-associated fibroblasts that build the desmoplastic stromal barrier around pancreatic tumors,” explained Pure and Bajbouj.
The team believes hitting the stromal barrier could reveal the tumor to the immune system to promote anti-tumor activity.
“This approach targets the barrier-building fibroblasts – not the cancer cells themselves.”
Homing in
The research team out of Penn set about building an in vivo platform from scratch. To deliver their mRNA directly to pancreatic tumors, they created what they call tLNPs: LNPs targeted to T cells.
“The tLNP platform enables in vivo engineering of T cells – reprogramming them directly within the host without ex vivo cell isolation or manipulation,” said Pure and Bajbouj. “The LNP is coated with an antibody that binds to T cells. After systemic administration, tLNPs deliver the mRNA into T cells, which then generates FAP-attacking CAR T cells that kill cancer-associated fibroblasts and disrupt the extracellular matrix.”
Traditional CAR T therapy requires weeks of lab manufacturing, and the efficiency can be quite variable.
Bajbouj, Pure, and their team coated their tLNPs in anti-CD5 antibodies to target T cells to reduce the likelihood of off-target effects. Their approach mirrors strategies BioXconomy has reported on in the past to deliver mRNA to immune cells and other tissues.
“Anti-CD5 antibodies conjugated to the tLNP surface bind predominately to T cells, which are characterized by high surface density of CD5,” noted Bajbouj and Pure. “This targeted delivery enables preferential uptake of mRNA encoding a CAR directed against FAP.”
The Penn team then turned to animal models to see if their approach could indeed generate CAR T cells in vivo. Helping a patient mount an immune response against the stromal barrier without requiring any ex vivo manipulation would represent a huge advantage for scaling CAR T therapies.
“Traditional CAR T therapy requires weeks of lab manufacturing, and the efficiency can be quite variable,” explained Pure and Bajbouj. “Manufacturing also requires highly specialized facilities and is quite costly relative to in vivo generation.”
They hope that this approach will make it easier to poke a hole in pancreatic cancer’s defenses. Moreover, their approach doesn’t use any viral vectors, which can come with their own safety concerns.
“Our FAP-CAR therapy is stable when frozen and can be readily shipped and administered elsewhere,” the pair said. “It does not require lymphodepletion as traditional CAR-T therapies do and is also potentially safer because mRNA is transient and does not integrate into the host genome.”
Breaking the siege
To test the efficacy of their in vivo CAR-T platform, the researchers separated pancreatic tumor-bearing mice into four groups. This included two control groups: one treated with saline and the other with the tLNP carrying a defunct mRNA sequence as a negative control. Among the two treatment groups, one was treated with the in vivo LNP-delivered FAP-mRNA therapy and the other with the ex vivo retrovirally transduced FAP CAR T cells to represent the current treatment standard.
The researchers found that the ex vivo therapy in the absence of lymphodepletion – an arduous process that can be harsh on patients – showed very small amounts of FAP CAR T cells expressed in the spleen, blood, or tumors. This result reflects the need for lymphodepletion in the clinic.
The in vivo mRNA strategy, however, showed robust expression three days after infusion, present in 45% of splenic CD3 T cells and almost 70% of peripheral blood T cells. More importantly, the FAP-targeting immune cells could be seen in more than 35% of T cells infiltrating the tumor.
“The tLNP platform rapidly yielded abundant CAR T cells amongst tumor-infiltrating T cells,” noted Bajbouj and Pure. “This is compared to less than 10% with conventional [ex vivo] CAR T cells at any given time point, even at the peak of their expansion.”
The in vivo generated CAR T cells were also able to generate similar expression levels comparable to ex vivo strategies when patients are “preconditioned” via immunosuppression.
“Certainly, a clinical advantage is that the FAP-CAR tLNP approach does not require patient preconditioning (lymphodepletion),” stated Pure and Bajbouj. Preconditioning can leave patients vulnerable to infection until their immune system is restored, a process that can take months.
Results also showed that in vivo generated FAP CAR T cells exhibited superior depletion of stromal cells within the tumor microenvironment compared to the traditional ex vivo strategy. Through flow cytometric analysis, Bajbouj and their team were able to record a 1.5-fold increase in invading immune cells within tumors versus control groups.
“The platform acts through two coordinated mechanisms,” commented Bajbouj and Pure. “First, tLNP-generated CAR T cells kill cancer-associated fibroblasts, causing the dense physical barrier (the stroma) to degrade. Second, this breakdown allows endogenous T cells to infiltrate the tumor.”
These findings were positively reflected in tumor cell growth. Ten days post a single infusion, in vivo-generated CAR T cells showed a 74% reduction in pancreatic tumor volume, compared to 48% achieved by the ex vivo therapy. In comparison, control groups experienced an almost 400% increase in tumor weights.
FAP-CAR T cells dismantle the stromal barrier and potentially enable the patient’s immune system to mount an effective immune response.
As an additional benefit, no systemic toxicity was seen in vivo-treated mice, and the therapy delivered consistent anti-tumor effects at either comparable or greater levels than traditional ex vivo therapies.
However, the researchers did note that, due to utilizing mRNA, the therapy is transient. Therefore, they were unable to detect almost any in vivo CAR T cells after the ten-day period. Bajbouj, Pure, and their team consider this to be a fair trade-off for not requiring traditional intensive immunosuppression, as is the case for ex vivo therapies, and being able to generate higher levels of T cells earlier in the process.
Long-term, the team hopes that these preclinical results will make the jump to human trials.
“FAP-CAR T cells dismantle the stromal barrier and potentially enable the patient’s immune system to mount an effective immune response,” Bajbouj and Pure said. “This could enhance the efficacy of combination therapies that previously failed due to the restrictive tumor microenvironment.”
Quotes have been lightly edited for clarity.
Header image depicts the pancreas and the pancreatic duct by Johann Georg Wirsung, c. 1644.