It’s not often that a failed clinical trial leads to a scientific breakthrough. But when patients in the U.K. started showing adverse side effects during a Phase II cancer immunotherapy trial, researchers at La Jolla Institute for Immunology (LJI) Center for Cancer Immunotherapy, and at the University of Liverpool, went back through the data and worked with patient samples to see what might have been the cause.

Their newly reported findings provide important clues to why many immunotherapies trigger dangerous side effects—immune-related adverse events (irAEs)—and the team’s subsequent experiments in mouse models point to a potentially better immunotherapy dosing strategy when treating patients with solid tumors, which might avoid these adverse side effects.

“This work shows the importance of learning from early-stage clinical trials,” said La Jolla Institute for Immunology professor Pandurangan Vijayanand, MD, PhD, who co-led the research with Christian H. Ottensmeier, MD, PhD, FRCP, a professor with the University of Liverpool, the Clatterbridge Cancer Centre NHS Foundation Trust, and adjunct professor at LJI. “This research illustrates how you can go from a clinical study to a mouse study to see what’s behind toxicity in these patients,” added LJI professor and CSO Mitchell Kronenberg, PhD, whose lab led much of the mouse model work for the new study.

The team reported its findings in Nature, in a paper titled, “Intermittent PI3Kδ inhibition sustains anti-tumor immunity and curbs irAEs.”

Phosphoinositide 3-kinase δ (PI3Kδ) has a key role in lymphocytes, and inhibitors that target this PI3K have been approved for treatment of B-cell malignancies, the authors explained. But while studies in mouse models of solid tumors have demonstrated that PI3Kδ inhibitors (PI3Kδi) can induce antitumor immunity, “… its effect on solid tumors in humans remains unclear.”

Both Vijayanand and Ottensmeier are physician-scientists, and Ottensmeier is an attending oncologist who treats solid tumor patients. In just the last decade, he has seen more and more patients thrive thanks to advances in immunotherapies, which work with the immune system to kill cancers.

“In the oncology world, immunotherapy has revolutionized the way we think about treatment,” sayidOttensmeier. “We can give immunotherapies to patients even with metastatic and spreading disease, and then just three years later wave goodbye and tell them their cancer is cured. This is an astounding change.”

Unfortunately, only around 20–30 % of patients with solid cancers who are given immunotherapies do then go into long-term remission. Some people see no change after immunotherapy, but others develop serious problems in their lungs, bowel, and even skin during treatment. These side effects can be debilitating or even fatal, and these patients are forced to stop receiving the immunotherapy.

PI3K inhibitors were initially considered to target mainly PI3K activity intrinsic to cancer cells, and this was the underlying rationale for testing inhibitors against the leukocyte-enriched PI3Kδ in B-cell malignancies, the authors continued. “However, subsequent studies have shown that PI3Kδ inhibition also has clear immunomodulatory activities, largely T cell-mediated, that were under-appreciated at the time of the early trials in B-cell malignancies, causing irAEs that have hampered clinical progress and utility.”

For their reported work, the researchers at LJI and the University of Liverpool worked with samples from a recent clinical trial in the U.K. involving patients with head and neck cancers, who were treated using an oral PI3Kδ inhibitor, AMG319.

PI3Kδ inhibitors are new to the cancer immunotherapy scene, but they hold promise for their ability to inhibit regulatory T cells (Tregs). Tregs normally try to stop other effector T cells from targeting the body’s own tissues. Inhibiting Tregs inside tumors could allow effector T cells to let loose and generate cancer-killing CD8+ T cells. “Having an oral tablet that can take off the brakes—the Tregs—can be a great asset for oncologists,” said Vijayanand.

At the time of the team’s trial, PI3Kδ inhibitors had proven effective for B-cell lymphomas but had not yet been tested in solid tumors. As the investigators acknowledged, while the evidence does suggest that PI3Kδi preferentially inhibit Tregs over other T cell subsets “… so far, no trials have been performed to explicitly explore this concept in humans.”

Their reported neoadjuvant, double-blind, placebo-controlled randomized Phase II trial evaluated AMG319 in human patients with head and neck cancer. Unfortunately, they found that more than half of patients given the drug had to discontinue treatment early because they developed inflammation in the colon, a condition called colitis. “At the tested doses of AMG319, immune-related adverse events (irAEs) required treatment to be discontinued in 12 out of 21 of patients treated with AMG319, suggestive of systemic effects on Treg cells,” the team noted. “Accordingly, in mouse models, PI3Kδi decreased the number of Treg cells systemically and caused colitis.”

“We thought this drug wouldn’t be toxic, so why was this happening?” asked Vijayanand. LJI instructor Simon Eschweiler, PhD, then spearheaded the effort to go back and see exactly how PI3Kδ inhibitor treatment affected immune cells in the patients. Using single-cell genomic sequencing, he showed that in the process of increasing tumor-fighting T cells in tumors, the PI3Kδ inhibitor also blocked a specific Treg cell subset from protecting the colon. Without Tregs on the job, pathogenic Th17 and Tc17 cells moved in and caused inflammation and colitis. “Single-cell RNA-sequencing analysis revealed a PI3Kδi-driven loss of tissue-resident colonic ST2 Treg cells, accompanied by expansion of pathogenic T helper 17 (TH17) and type 17 CD8+ T (TC17) cells, which probably contributed to toxicity; this points towards a specific mode of action for the emergence of irAEs,” the investigators further explained.

It was evident that the cancer trial patients had been given a larger PI3Kδ inhibitor dose than they needed, and the immunotherapy had thrown the delicate composition of immune cells in the gut out of balance. The team also found, through studies in mouse models, that intermittent dosing could be a valid treatment strategy that combines sustained antitumor immunity with reduced toxicity. They are now designing a human clinical trial to test the intermittent dosing strategy in humans.

“It is clear that in the neoadjuvant setting in patients with HNSCC, at the evaluated doses and with daily scheduling, PI3Kδ inhibition has an unfavorable safety profile, limiting its feasibility and clinical benefit by causing frequent and severe grade 3/4 irAEs, probably driven by modulation of Treg cell behavior in nonmalignant tissues,” they wrote. “A modified treatment regimen with intermittent dosing of PI3Kδi in mouse models led to a significant decrease in tumor growth without inducing pathogenic T cells in colonic tissue, indicating that alternative dosing regimens might limit toxicity … On the basis of our findings, decreased dosages or an altered PI3Kδi treatment regimen will be required in solid tumors—especially in immune-competent patients—in order to be able to exploit the clear anti-tumor immune response induced by PI3Kδi while limiting the adverse effects associated with reduced Treg function in healthy tissues.”

It’s also possible that the pathway leading to the toxicity seen in their reported study may be broadly applicable to other organs harboring similar Treg cells, and to other Treg cell-targeting immunotherapies like anti-CTLA-4, Eschweiler suggested.

But how might the lack of toxicity in trials for B-cell lymphomas be explained? Eschweiler suggests that lymphoma patients in previous studies had been given several prior therapies leading to an overall immunocompromised state. This means the lymphoma patients didn’t have the same type—or the same magnitude—of immune response upon PI3Kδ inhibition. In contrast, the head and neck cancer patients were treatment-naive. Their immune system wasn’t compromised, so the immune-related adverse events were both more rapid and more pronounced.

Overall, the new study shows the importance of studying not just personalized therapies but personalized therapy doses and schedules. As Ottensmeier explained, doctors ten years ago only had one type of immunotherapy to offer. It either helped a patient or it didn’t. Doctors today have a rapidly growing library of immunotherapies to choose from. So in the context of their reported findings, the team stated, “… our data suggest that the unique cellular composition of effector versus regulatory cells in the TME of each patient might be an important determinant of the efficacy of PI3Kδ inhibition. Thus, PI3Kδi might be especially useful in patients with high levels of intratumoral Treg cells and an unfavorable ratio of Treg versus CD8+ tumor-infiltrating lymphocytes (TILs) in pre-treatment samples.”

Vijayanand and Ottensmeier are among the first researchers to use single-cell genomic sequencing tools to determine which therapeutic combinations are most effective in individual patients—and the best timeline for giving these therapies. In a 2021 Nature Immunology study, the pair showed the potential importance of giving immunotherapies in a specific sequence. “If you design your clinical trials well and apply sophisticated genomics, you have a lot to learn,” said Vijayanand. “You can figure out what’s happening and go back to the patients.”