Drugs that enhance the body’s immune response against cancer may also be altering one of its most tightly guarded boundaries: the blood-brain barrier (BBB).
A recent study published by Yuval Shaked at the Technion-Israel Institute of Technology and his team, in Cancer Discovery, finds that PD-1 inhibitors, a widely used class of cancer immunotherapy, can prompt immune cells to produce a protein that makes the barrier more permeable. This could potentially change how cancer and its treatments affect the brain.
Many conventional anti-cancer drugs cannot cross the BBB, which is a tightly packed lining of cells that controls what passes from the bloodstream into brain tissue, limiting their effectiveness against brain tumours. So the brain was long thought to be largely insulated from the immune system, but growing evidence shows it can mount meaningful immune responses. In this context, immunotherapy works by activating circulating immune cells that can cross the BBB and target tumour cells within the brain.
A type of immunotherapy called immune checkpoint inhibitors (ICIs) block signals that prevent immune cells from attacking tumours, allowing the body’s natural defences to respond more strongly. While ICIs have been shown to reduce tumour burden within the brain, responses among patients with brain metastases vary and the reasons remain unclear.
“Our work focuses on understanding how cancer treatments affect the body, not just the tumour. In some cases, therapies can trigger responses in normal host cells, such as immune cells, that unintentionally make the environment more favourable for cancer growth,” Abhilash Deo, a postdoctoral researcher at the Shaked Lab and the study’s lead author, said.
Brain environment
To understand how immunotherapy affects the brain immune environment, researchers examined brain tissue from mice with breast tumors treated with anti-PD-1 therapy. They observed a loss of cells that maintain blood vessel stability, weaker barrier proteins, and higher immune cell entry into the brain, suggesting that the BBB was becoming leaky.
Mice treated with anti-PD-1 also showed an increase in brain metastases, likely due to the compromised barrier. Notably, these effects were observed only with anti-PD-1 and not with other ICIs, highlighting a unique host response induced by treatment.
“Our data show that anti-PD-1 therapy can boost anti-tumour immunity in the brain, but in resistant cancers, it may also increase metastasis by altering the host immune environment,” Dr. Deo said. “This could help explain why patients with brain metastases show varied responses to immunotherapy.”
According to Nirmal Raut, medical oncologist at the Bhaktivedanta Hospital and Research Institute in Thane, treatment responses to ICIs vary widely among patients with brain metastases, ranging from complete remission to rapid disease progression (seen in about 20% of cases after treatment starts).
“We often see discordant responses, where disease outside the brain is controlled, but new brain lesions appear, or vice versa, suggesting that the brain-immune ecosystem is distinct from the rest of the body,” he added.
Dr. Raut noted that even when a tumour responds to treatment in organs such as the lung or liver, the BBB can act as a sanctuary where sub-therapeutic drug levels allow cancer cells to survive and evolve.
Key mediator
When untreated animals were injected with plasma from mice treated with anti-PD-1, the researchers observed BBB leaks, suggesting that treatment-induced ICIs were disrupting the barrier. Comparing plasma protein profiles of treated and untreated animals, the team identified several proteins linked to BBB disruption. Among these, removing a protein called DKK1 reduced BBB leakage.
Importantly, these findings were reflected in patient data. MRI scans of patients with lung cancer who received anti-PD-1 therapy showed increased cancer spread within the brain. Higher levels of plasma DKK1 were also linked to a greater occurrence of brain metastases and a shorter period before the disease worsened, particularly among patients who responded poorly to treatment.
“This aligns with the idea that elevated DKK1 may point to a more permissive brain environment for metastasis,”Dr. Raut said
He added that the increased contrast seen on some MRI scans after starting immunotherapy may not always indicate “pseudoprogression” or inflammation but could instead reflect actual BBB leakage caused by activated immune cells.
Double-edged role
Chakor Vora, a medical oncologist at the Renatus Cancer Centre, Pune, explained that most chemotherapeutic drugs can’t cross the BBB, posing a major challenge in treating brain metastases.
Opening of the BBB following anti-PD-1 therapy may therefore improve their delivery to the brain. The researchers found that anti-PD-1 therapy followed by cisplatin chemotherapy improved survival in mice with brain metastases as well as increased drug accumulation in the brain, highlighting a double-edged role.
Dr. Raut said that among patients who don’t respond to treatment, opening the BBB using anti-PD-1 therapy could inadvertently allow circulating cancer cells also to enter the brain, potentially increasing the risk of new metastases.
“However, for patients with resistant disease, this same vulnerability could be exploited to improve drug delivery to the brain,” he said.
Rahul Solanki, a medical oncologist and a physician of nuclear medicine in Adelaide in Australia, added that once cancer has spread to the brain, the BBB may already be disrupted, and such patients are often excluded from clinical trials. Since medical workers can’t measure drug levels in the brain, DKK1 may be a promising biomarker that could help identify patients at higher risk of developing brain metastases during treatment.
“Patients with advanced cancer but without active brain metastases would be better candidates to understand how anti-PD-1 therapy affects treatment response and the risk of metastasis,” Dr. Solanki said.
“We typically use a combination of chemotherapy and immunotherapy in high-risk patients with brain metastases, who test positive for immune biomarkers. However, these findings need to be validated in larger studies involving human patients,” Dr. Vora stressed.
“If these findings are confirmed in larger human trials, they could change how we sequence treatment,” Dr. Raut added.
Shweta Yogi is a freelance science writer.