“This study provides mechanistic insight into a phenomenon that many melanoma experts have observed firsthand in the clinic and that has recently been described in retrospective studies: poor response to immunotherapy following the development of resistance to [targeted] therapy,” Brent Hanks, a medical oncologist at Duke University who was not involved in this study, tells The Scientist in an email. “Overall, this is an important contribution to melanoma research that may have implications in the management of other . . . cancers as well.”
Indeed, it was early clinical data that sparked the interest of Anna Obenauf, a cancer researcher at the Research Institute of Molecular Biology in Vienna, Austria, who led the international team behind the new study. “This is a clinical puzzle in a way, because how can these two different types of therapies be connected to each other, and this resistance to one lead to cross-resistance to the other?” While targeted therapy blocks specific molecular pathways within cancer cells to stop proliferation, immunotherapy works by stimulating immune cells to eradicate tumor cells.
Their work showing that you can reverse the phenotype by adding in these dendritic cell–stimulating agents was a nice proof of principle to show that it really was those cells that were being restricted.
—Brian Ruffell, Moffitt Cancer Center
Obenauf and her colleagues started by recapitulating these clinical observations in a mouse model. Using two murine melanoma cell lines, the researchers established tumors in mice, which they treated with dabrafenib, a targeted therapy approved for use in the treatment of melanoma patients who have a mutation in the BRAF gene. Dabrafenib interrupts the MAP kinase pathway by inhibiting the B-Raf enzyme. While the tumors initially responded to the therapy, the cancer eventually relapsed and became resistant. Taking cells from the treatment-sensitive tumors and the treatment-resistant tumors, the researchers established cell lines. These cells were again injected into mice, which were treated with anti-PD-1 or anti-CTL-4 checkpoint inhibitors, immunotherapies aimed at releasing the brake on the immune system. Anti-PD-1 and anti-CTL-4 checkpoint inhibitors are also approved for treating certain patients with melanoma.
Using this approach, the researchers could implant resistant tumors into mice that had not been exposed first to the targeted therapy. This allowed the team to assess whether the targeted therapy has a direct effect on immune cells that could lead to immunotherapy resistance, or if something else is going on within the tumor. It turned out to be the latter. “[Treatment-resistant] tumors are indeed cross-resistant to checkpoint inhibitors,” says Obenauf.
Immunotherapies usually act by promoting T cell responses, so the group looked more closely at how the mice’s T cells behaved. While T cells were able to kill treatment-resistant tumor cells in vitro, when the researchers used a mouse model lacking endogenous T cells and added T cells they could track using luciferase, they saw that the T cells couldn’t infiltrate the resistant tumor; the tumor kept growing. “That has led us to the question [of] whether the tumor microenvironment is mediating resistance,” Obenauf recalls.
So the researchers created mix-and-match melanoma mice. When they placed treatment-resistant tumor cells within a large treatment-sensitive tumor, the resistant tumor cells were killed. It seemed that treatment-sensitive tumors had “an immune-permissive tumor microenvironment,” Obenauf explains. Conversely, when the researchers placed treatment-sensitive cells within a large treatment-resistant tumor, the cells survived, apparently shielded from T cell–mediated killing.
A tumor naive to targeted therapy (top) contains many more immune cells (red and green) than one that has acquired resistance (bottom).
RNA sequencing and flow cytometry analysis revealed that dendritic cells, a cell type crucial for activating the immune system during immunotherapy, were less abundant in mice with treatment-resistant tumors. When the researchers co-cultured dendritic cells with T cells, they saw that the dendritic cells from resistant tumors didn’t activate T cells or spur them to proliferate as dendritic cells from sensitive tumors did. Collaborating with a team at the University of Sydney in Australia, the group acquired biopsies from patients with melanoma who were treated with a targeted therapy. Once the patients had become resistant to the treatment, their tumors contained fewer dendritic cells than before.
Collectively, the results suggest that a drop in dendritic cells “generates an immune-evasive tumor microenvironment that is poorly responsive to subsequent checkpoint inhibitor immunotherapy,” Hanks explains.
Notably, the effect was reversible. After treating the mouse models with experimental immunostimulants that mature and expand dendritic cell pools, the researchers saw greater numbers of T cells infiltrating the animals’ tumors, which shrunk as a result. “Their work showing that you can reverse the phenotype by adding in these dendritic cell–stimulating agents was a nice proof of principle to show that it really was those cells that were being restricted,” Brian Ruffell, a cancer immunologist at the Moffitt Cancer Center who was not involved in this study, tells The Scientist.
“I think this [study] really breaks down some of the biology of why you’d want to treat patients with immunotherapy before you come in and allow resistant clones to develop from targeted therapy,” Ruffell adds. “From a basic science point of view, it really helps to add to the growing body of literature that we need to study all therapies in the context of immunotherapy or the immune system.”
To understand how cells were developing cross-resistance, Obenauf and colleagues analyzed the transcriptomes of cells from both mice and patient samples that had grown resistant to a therapy that targeted the MAP kinase signaling pathway. A hyperactive MAP kinase pathway leads to uncontrolled cell proliferation but is turned down by the inhibitor, and tumors shrink in response. When tumors relapse and become resistant to inhibitors, the MAP kinase pathway is frequently re-activated.
We can very strongly conclude that that pathway reactivation is what’s driving the immune therapy resistance.
—Kristian Hargadon, Hampden-Sydney College
In their samples, Obenauf’s team identified a signature of genes that are differentially expressed in targeted therapy–resistant tumors versus sensitive tumors. Using a computational analysis to find the regulators that govern this genetic signature, the scientists found that the MAP kinase signaling pathway was turned on again in resistant tumors and apparently now driving immune evasion. “It was surprising that the differences between the [treatment-sensitive] and the [treatment-resistant] tumors were predicted to be driven by the MAP [kinase] pathway,” Obenauf says, “because the [treatment sensitive] tumors, despite the MAPK pathway being already hyperactive, were so sensitive to immunotherapy, whereas the [treatment resistant] tumors, where the MAPK pathway is being re-activated, were so resistant to immunotherapy.”
It turned out that the MAP kinase pathway in resistant tumors more strongly drives gene expression of target genes than it does in sensitive tumors. Components of the pathway also had access to new gene regulatory sites, meaning that they could drive the expression of different genes. “The MAP kinase pathway, the same pathway that is very important for tumor initiation, is rewired and enhanced in this process of therapy resistance to establish a very different immune phenotype,” says Obenauf.
“We can very strongly conclude that that pathway reactivation is what’s driving the immune therapy resistance,” says Kristian Hargadon, a biologist at Hampden-Sydney College not connected with the study. “And that is something that people would not have expected up until this point, yet now that explains a lot of previous observations.”
Pulling all these strands together, the team treated mice that had targeted treatment–resistant tumors with a MEK inhibitor, which inhibits the MAP kinase pathway at a different point than does the targeted treatment used initially. In vitro, this inhibition reverted the expression of 80 percent of the genes that formed the signature for resistance back to the treatment-sensitive expression signature. When mice with a treatment-resistant tumor were given the MEK inhibitor, dendritic cells became more numerous and active, inducing T cell proliferation. When the researchers gave the animals immunotherapy, the T cells were able to bring the tumors under control, and the mice survived longer. “The effects were quite drastic, indicating that the MAP kinase pathway along with the dendritic cells really are responsible for mediating cross-resistance,” says Obenauf.
“This is a very elegant, intricate, thorough study,” Hargadon concludes. “Several different tumor models were studied, several different therapeutic regimens were evaluated, all pointing to the same phenomenon here.”