CAR T cells were engineered to resist liver cancer’s stop signal

CAR T cells were engineered to resist liver cancer’s stop signal

A solid tumour can display a convenient target to immune cells while simultaneously telling them to stand down. That contradiction is one reason CAR T cells have struggled to control liver cancer.

CAR T therapy begins with a patient’s own T cells. The cells are collected, genetically modified to recognise a molecule on cancer cells, expanded and returned to the patient. The approach has produced strong results in several blood cancers, but engineered cells often lose activity inside the dense and suppressive environment of a solid tumour.

One of the molecular brakes in that environment is TGFβ. This signalling protein can bind to receptors on T cells and suppress their proliferation and antitumour functions.

The experimental treatment C-CAR031 modified the cells in two complementary ways. Its CAR receptor recognised glypican-3, or GPC3, a surface protein frequently expressed by hepatocellular carcinoma. The cells also carried a shortened TGFβ receptor that could interfere with the suppressive signal without transmitting the usual inhibitory instruction into the cell. In effect, the CAR T cells were given partial protection against one of the tumour’s immune defences.

The first-in-human phase I trial enrolled 36 people with advanced liver cancer that had continued to progress after previous systemic treatment. Following lymphodepleting chemotherapy, each participant received a single infusion of their own engineered cells.

Some reduction in target tumour lesions was recorded in 32 patients. Any measurable shrinkage, however, is not the same as a formal clinical response. Sixteen of the 36 participants, or 44.4%, met the established criteria for an objective response. The median best reduction in target lesions was 41.6%.

The responses were often short-lived. Their median duration was 4.4 months, while median progression-free survival was 4.2 months. Median overall survival was 14.2 months, but a small trial without a control group cannot determine whether the treatment extended survival compared with existing therapies.

Cytokine release syndrome occurred in 34 of the 36 participants. This systemic inflammatory reaction was mostly mild or moderate, although two patients developed grade 3 events. Nine participants experienced grade 3 or higher non-haematological adverse events.

Tumour samples offered clues to resistance. Some cancer cells reduced or completely lost GPC3, leaving the CAR T cells without the target they had been engineered to recognise. Other tumours retained strong TGFβ signalling. The modified receptor reduced the cells’ sensitivity to this pathway, but it did not eliminate the rest of the suppressive tumour environment.

This was an early, small and non-randomised study. Participants were selected for GPC3-positive tumours, and many screened patients did not meet that requirement. The efficacy figures therefore apply to a narrowly selected population and need confirmation in larger comparative trials.

For CAR T cells facing a solid tumour, recognising the right target is only half the task. The cells must also remain functional in a biological environment built to restrain them.