Avatar Dragons

Until recently, the way we’ve been treating cancer in modern oncology has been a matter of direct confrontation. You had surgery to take out a tumor, radiation to do some local damage, and chemo to put the whole body’s fast-dividing cells in their place. They are still essential, but they have a flaw in common: they tend to view cancer as something to be obliterated, not as a living thing that can adapt, go to ground, and put up a fight.

Then came immunotherapy and it put a different spin on things. The question is no longer just how to put an end to cancer from the outside, but how to put the body’s own defenses to work for us. If you look at the literature, one thing stands out: what was once a type of treatment is now shaping up to be a discipline of its own.

Take immune checkpoint inhibition, for instance. That’s where you use drugs to put a stop to the ways a tumor will try to dull a T-cell’s response. We’ve seen this with inhibitors for CTLA-4, PD-1 and PD-L1, and it has made a world of difference for patients with melanoma, lung or kidney cancer. What you get with these is longevity in the response — something you don’t often see with the old guard of oncology.

It’s not all good news, though. Checkpoint inhibitors are a case study in why you can’t have a one-size-fits-all. Some tumors are “cold” and won’t budge. Some give in and then build up a resistance. And if you let off the brakes on the immune system, you run the risk of it lashing out at healthy tissue.

On the other side of the coin is adoptive cell therapy. Here, you can actually take a patient’s immune cells out, work on them, and put them back in. CAR-T is the poster child for this, with some very impressive showings in blood cancers. And in 2024 we hit a mark with TIL therapy: lifileucel made history as the first cellular therapy to get the FDA’s stamp of approval for a solid tumor.

Then there’s the more tailored approach. You have bispecifics that put an immune cell and a tumor in physical contact, or vaccines made to order for a patient’s specific neoantigens. Even oncolytic viruses have a role to play in making a tumor a target for the immune system.

Put it all together and the trend is obvious. We are done with the blunt instrument of general immune activation. We’re in the business of fine-tuning. It’s about making the immune system more discerning, so cancer has a harder time getting away with it.

If you look at where we are with immunotherapy, the big story is that it’s no longer just about checkpoint inhibitors. While those are still the bedrock of the field, we’re seeing a lot more emphasis on putting things together: combinations, engineered cells, bispecifics, and even reworking the tumor microenvironment.

Then there’s CAR-T. It’s hard to overstate what an achievement it has been for certain blood cancers like large B-cell lymphoma or multiple myeloma. If you follow the data, you see patients who have been in remission for the long haul after treatment. But let’s be honest, as the literature will tell you, it hasn’t put an end to cancer. You don’t see the same kind of results in solid tumors.

And for good reason. A solid tumor is a tough nut to crack. It’s hard for immune cells to get in, and once they do, the environment is hostile. A CAR-T might be set up to target one thing, but a solid tumor is made of all kinds of different cell types; some won’t have the right marker, and you get antigen escape. Add in the low oxygen, the weird vasculature and the metabolic strain, and you have a place where your T cells can get worn down.

That’s the impetus for the more intricate cell therapies we’re working on now. We’re looking at multi-target and “armored” CARs, TCR-engineered T cells, and allogeneic options. An off-the-shelf product would be a boon for cost and time, of course, but you have to deal with the risk of rejection or graft-versus-host disease.

Bispecifics are a step in that direction too. They’re a lot more straightforward than having to harvest and build a patient’s own cells. You have one arm latching onto a tumor and the other on a T cell’s CD3, and you put them in each other’s faces. It’s more of a scalable approach, and while they’re doing well in hematologic malignancies, the solid tumor wall is still there.

We’re also seeing a come-back of the cancer vaccine, in particular the personalized kind. The KEYNOTE-942 trial was telling: in high-risk melanoma, an mRNA neoantigen therapy with pembrolizumab did better than the drug by itself. It’s a sign that mRNA can do more than just ward off viruses; it can be used to put a patient’s own immune system to work.

On the other end of the spectrum you have oncolytic viruses. They’ll go in and take out a tumor cell, and in the process put out some antigens and inflammatory markers. In an ideal world, you can make a cold tumor hot. Down the line, I think you’ll see them used most effectively when you pair them with a checkpoint or a vaccine.

Up until recently, the way we’ve been treating cancer in modern oncology has been a matter of direct confrontation. You had surgery to take out the tumor, radiation to do some local damage, and chemo to put down any fast-dividing cells in the body. They are still essential, but they have a flaw in common: they tend to view cancer as something to be obliterated, not as a living thing that can adapt, lie low, and put up a fight.

Immunotherapy has put a new spin on things. The question is no longer just how to put an end to cancer from the outside, but how to put the body’s own defenses to work for you — whether by training them or rethinking how they’re put together. If there’s one thing that comes through in the literature, it’s this: what we used to call a type of treatment is now an engineering discipline.

Take immune checkpoint inhibition, for instance. It’s the first of the big three. Inhibitors like those for CTLA-4, PD-1 and PD-L1 put a stop to the ways a tumor would normally dull a T-cell’s response. We’ve seen them make a world of difference in melanoma, lung and kidney cancers. What you get with these is longevity; in some cases, the effect holds for years, and that doesn’t happen with the old guard of oncology.

There’s a cautionary tale to it, though. Not every patient will see a benefit. Some tumors are “cold” and don’t have the kind of immune presence needed to make a dent. Or a patient might have an initial win, only for the cancer to find a way around it. And then there are the side effects — when you let off the brakes on the immune system, it can sometimes turn on healthy tissue.

Then you have adoptive cell therapy. Here, you don’t just nudge the immune system in situ; you can actually pull cells out, do some work on them, and put them back in. CAR-T is the poster child for this, with some of the best outcomes in B-cell cancers and myeloma. And in 2024 we hit another mark with TIL therapy, when lifileucel was given the green light by the FDA as the first cellular option for a solid tumor.

The third leg of the stool is about redirection and making things personal. With bispecifics you can put an immune cell in physical contact with a tumor. A custom vaccine can be made to zero in on the neoantigens of a particular patient’s disease. Even oncolytic viruses have a role to play, infecting the tumor and making it a target for the immune system.

Where is all this heading? Toward a more exact science of the immune system. We aren’t just looking to beef up the body’s response anymore. We want it to be more discerning, and we want to make sure cancer can’t slip away.

You won’t find a lack of potency in cancer immunotherapy. The issue is that it’s not always there when you need it. You have some patients with a given cancer who have a remarkable, enduring response, and then you have others with the very same diagnosis for whom the treatment does nothing. That kind of variability comes up time and again in the literature. It’s a powerful tool, to be sure, but one that is hard to put your finger on, hard to manage, and hard to make work on a larger scale.

Tumor heterogeneity is at the root of it. A tumor isn’t made up of a uniform set of cells. Even in one mass, you can have cells with different mutations and antigens that will react to therapy in their own way. Hit them with an immunotherapy aimed at a particular antigen and the ones without it are left to live. They’ll multiply and you get a relapse. We call it antigen escape, and it’s a hurdle for everything from CAR-T to bispecifics and vaccines.

It’s a tougher proposition with solid tumors. With blood cancers, the immune system has a clear line of sight to the problem in the marrow or lymph. Solid tumors are more like a walled city. There’s the density of the tissue, the odd blood flow, the pressure, and a host of other factors. For a treatment to do its job, the immune cells have to make it in, put down roots, and be in a position to take out the cancer.

If you read any of the new research, the tumor microenvironment is what you see over and over. Tumors don’t just sit there; they put up a defense. They’ll bring in regulatory T cells and macrophages to stifle the T-cells, or put out chemical signals to put the immune system on the back foot. In a sense, the cancer is reconfiguring the terrain.

And that puts a cap on what checkpoint inhibitors can do. Take away the brakes with a PD-1 or CTLA-4 drug and you might be fine if there are already troops in the area. But a “cold” tumor with little to no immune presence is another story. If the right cells aren’t there, there’s only so much a blockade can do.

You can see the next chapter of cancer immunotherapy in three broad changes: we’re going to see a move from one-off therapies to combination regimens, from a sledgehammer approach to immune activation to something far more exact, and from treating what’s in front of us to a more adaptive, made-to-order kind of care.

For one, checkpoint inhibitors are not going anywhere. They are the bedrock of the field for good reason; PD-1, PD-L1 and CTLA-4 blockade showed us that you could get the immune system to put up a long-lasting fight against a tumor. But on their own, they don’t have the answer for cold tumors or when there is too much heterogeneity at play. I’d expect to see them used as part of a bigger plan, where you first put a target on the back of the tumor and then let the immune cells do their work.

Then there are the personalized vaccines, which will be worth keeping an eye on. The case for them is hard to argue with: no two tumors are alike, and some of those unique mutations make for neoantigens the body can spot. We have the tools now — quicker sequencing, machine learning to call out antigens, and mRNA tech — to build a vaccine for the individual. We’ve already seen some of this in melanoma, where a custom neoantigen therapy was paired with a checkpoint inhibitor to good effect. It’s moving from being a nice theory to a real option.

That said, it all comes down to how well we can predict the right antigen. We still have to get better at figuring out which neoantigens a T cell will actually see. Down the line, you may find that a vaccine is put together using everything from HLA typing and spatial transcriptomics to AI to zero in on the best ones.

And as for cell therapy, we are due for an upgrade from the CAR-T of today. What we have is potent, but in a way, it’s a bit of a blunt instrument. It latches on to one thing and goes to work. The new wave of these therapies will be more nuanced. Think of cells with several receptors, some built-in logic, and even a safety valve so a physician can put the brakes on if there’s any trouble.

Synthetic biology is set to be the driving force here. You could have an immune cell that only fires when it sees two different cancer markers, sparing the healthy tissue. One that can hold its ground against a tumor’s attempts to put it to sleep, or even one that is told to self-terminate if it gets out of hand. In short, we are going from engineering a cell to recognize a target to engineering it to make decisions. You can count on off-the-shelf therapies to be a top priority for the industry, which is our fourth forecast. There’s no doubt that personalized cell therapies have their place, but they are costly, hard to put together and slow to come by. An allogeneic approach, using donor cells, lets you make them in bulk and have them on hand. Then there are CAR-NK cells; since natural killer cells don’t have the same kind of antigen-specific limitations as T cells and pose less of a graft-versus-host risk, they’re set to be a big deal. Of course, you still have to get around issues like rejection, safety and how to make them in the first place.

For the fifth one, we see in vivo engineering as a potential game-changer down the road. Why not skip the whole process of taking cells out, modifying them and putting them back in? You could just give the patient the genetic code and let the immune system do the work in situ. Do it right and you cut down on time and expense, opening up cell therapy to more people. The trick will be in the delivery and making sure you only edit the cells you mean to.

As for the sixth, biomarkers are going to be every bit as vital as the drugs. A good treatment is useless if you can’t tell who it’s for. We’ll be looking at all sorts of data points to make those calls: tumor mutational burden, neoantigens, what’s in the microbiome, even spatial maps of the immune system. AI will be there to make sense of it all.

If there’s one thing to remember, it’s this: “boosting the immune system” is an oversimplification. What we’re really doing with modern immunotherapy is dictating the terms — what the immune cells see, where they head, when they fire and when they stand down.

And don’t forget that cancer is an opponent that evolves. Put some pressure on it with a therapy and it will find a way to adapt. Go after one antigen and you might leave the negative ones behind. Block a checkpoint and another pathway will take over. So our approach has to be as fluid as the disease.

Which brings us to the fact that the most effective options are going to be in combination. Whether it’s a checkpoint inhibitor, a vaccine or a bispecific antibody, these are all tools in the same box.

In the end, we’re in the engineering phase of cancer immunotherapy now. We’ve made our case that the body can put up a fight and win, and for some, those wins are long-lasting. But the job is to make that the rule rather than the exception. Cancer is in the business of evading and stalling; we are trying to put an end to that. The therapies of tomorrow won’t be the ones that are just more of a brute force, but the ones that are exact.

If you would like to have a more immersive way to comprehend this paper, then check out this video on the subject.

Sources

1. Zhang, M., et al. “Advances in Cancer Immunotherapy: Historical Perspectives, Current Developments, and Future Directions.” Molecular Cancer, 2025.
2. Nakayama, I. “The Current Status of Immunotherapy and Future Horizon in Cancer Therapy.” Japanese Journal of Clinical Oncology, 2024.
3. Devaraji, M., et al. “Immune-Based Cancer Therapies: Mechanistic Insights, Clinical Progress, and Future Directions.” Journal of the Egyptian National Cancer Institute, 2025.
4. Zugasti, I., et al. “CAR-T Cell Therapy for Cancer: Current Challenges and Future Directions.” Signal Transduction and Targeted Therapy, 2025.
5. Trabolsi, A., et al. “Bispecific Antibodies and CAR-T Cells: Dueling Immunotherapies for Large B-Cell Lymphomas.” Blood Cancer Journal, 2024.
6. Weber, J. S., et al. “Individualised Neoantigen Therapy mRNA-4157 (V940) Plus Pembrolizumab Versus Pembrolizumab Monotherapy in Resected Melanoma: A Randomised, Phase 2b Study.” The Lancet, 2024.
7. Hu, L., et al. “FDA Approval Summary: Lifileucel for Unresectable or Metastatic Melanoma.” Clinical Cancer Research, 2025.
8. Qin, X., et al. “T-Cell Redirecting Bispecific Antibodies in Cancer Therapy.” 2024.
9. Xiao, D., et al. “Oncolytic Viruses: Advanced Strategies in Cancer Therapy.” Signal Transduction and Targeted Therapy, 2026.
10. Yang, K., et al. “Adaptive Resistance in Cancer Immunotherapy.” Cellular & Molecular Immunology, 2026.