The T Cell Fountain of Youth
Rewriting the rules of tumor control by preserving the immune reservoir
TL;DR Progenitor CD8 T cells, the stem-like precursors of exhausted T cells, offer a renewable, adaptable anti-tumor immune force, whereas terminally exhausted T cells are a dead end. In cancer immunotherapy, keeping T cells young and responsive beats trying to revive those that are irreversibly burned out.
Key points:
Terminal Exhaustion = Diminished Returns: Terminally exhausted CD8⁺ T cells express many inhibitory receptors and can no longer proliferate or kill effectively. They are an end-stage state with fixed epigenetic “scars” that even after curing the tumor remain largely irreversible. In other words, once CD8 T cells terminally differentiate into exhausted cells, their functional decline is largely locked in place.
Progenitor T Cells = Renewable Fighters: A subset of CD8⁺ T cells in tumors retain a progenitor-like, stem-cell quality (marked by TCF1) that lets them self-renew and spawn new killer T cells. These TCF1⁺ “stem-like” T cells have lower immediate cytotoxicity but high proliferative potential, and they are the only ones that proliferate and respond robustly to checkpoint inhibitors like anti-PD-1. They serve as a renewable reservoir that can continuously supply fresh effector T cells when properly stimulated.
Better Immunotherapy Outcomes: Patients whose tumors harbor more of these progenitor-like T cells tend to respond better to immunotherapy. In melanoma, for example, tumors enriched in TCF1⁺ CD8 T cells were correlated with superior responses to PD-1 checkpoint blockade and longer survival. By contrast, having only terminally exhausted T cells (without progenitors to rejuvenate the army) is associated with poor outcomes.
Therapies to Expand Progenitors: New immunotherapies aim to sustain or amplify the progenitor T cell pool. IL-15 agonists (e.g. N-803, an IL-15 superagonist recently approved for bladder cancer) preferentially expand stem-like CD8 T cells and memory subsets. Co-stimulatory signals (such as CD28/4-1BB agonism alongside TCR activation) and T cell help from CD4⁺ T cells and dendritic cells prevent early exhaustion. These interventions keep the T cell response going long enough to generate epitope spreading, the process by which the immune system broadens its attack to new tumor antigens, cutting off tumor escape routes.
Pan-Cancer Impact, Durable Immunity Over One-Off Hits: Focusing on progenitor T cells aligns with emerging standards of durable cancer immunity. In checkpoint inhibitor therapy, it means more T cells that can be reinvigorated again and again (rather than one-and-done effectors). In T cell engager therapies, it means designing treatments that provide the full activation circuit (TCR + co-stimulation + cytokines) so T cells don’t flame out after the first burst. And in the tumor microenvironment, it means reshaping conditions (e.g. reducing hypoxia and Tregs) to favor long-lived T cell responses over terminal differentiation. The result is immunotherapies that work smarter and last longer, across cancer types.
From One-Off Wins to Durable Immunity
Immunotherapy has transformed cancer care, but the next frontier is making responses last. Today’s biotech investors and startup teams are not just looking for therapies that can shrink tumors briefly, they want durable, long-term remissions. One key insight leading this paradigm shift: the quality and phenotype of T cells we target may matter as much as the quantity. Specifically, there’s growing evidence that targeting stem-like progenitor CD8⁺ T cells (and preserving their pool) is far more effective for lasting cancer immunity than trying to rely on terminally exhausted CD8⁺ T cells that have hit a biological dead end.
In essence, not all tumor-infiltrating T cells are equal. Some are “serial killers” in training, they can multiply and adapt, while others are more like spent ammunition, present in number but low in firepower and unable to self-renew. This report delves into the biology behind these differences and explains why therapeutics that nurture the renewable subset of T cells stand to deliver superior, durable outcomes across many cancer types (a true pan-cancer opportunity). We’ll start with a quick primer on T cell differentiation and exhaustion to set the stage.
T Cell Differentiation 101
CD8⁺ T cells (the cytotoxic T cells that kill infected or cancerous cells) don’t start out exhausted, they are made capable of robust responses under the right conditions. A naive CD8⁺ T cell, upon recognizing a tumor antigen (or any antigen) with sufficient stimulation, will proliferate and differentiate into an effector T cell. Effector CD8 T cells are like foot soldiers: they produce cytokines (e.g. IFN-γ) and cytotoxic enzymes (granzymes, perforin) to kill target cells. In acute infections (or perhaps an immunogenic tumor that is quickly cleared), some of these effectors transition into memory T cells once the antigen is gone, long-lived cells that can rapidly respond if the antigen reappears.
Cancer, however, is more akin to a chronic infection, the antigen (tumor neoantigens) persist over time. Facing continuous stimulation in the tumor microenvironment, CD8 T cells don’t form healthy memory; instead they often undergo T cell exhaustion, a differentiation state first observed in chronic viral infections and now recognized in tumors. Exhausted T cells (often noted as Tex) are dysfunctional: they upregulate multiple inhibitory receptors like PD-1, TIM-3, LAG-3, TIGIT, etc., lose their robust cytotoxicity and cytokine production, and stop proliferating well. This exhaustion is thought to be a protective mechanism to avoid overactivation and immunopathology under constant antigen drive, but in cancer it becomes a barrier to effective tumor clearance.
Importantly, exhaustion isn’t a single uniform state, it’s a spectrum or hierarchy of differentiation. Researchers have identified two major subsets in exhausted CD8 populations within tumors:
Progenitor exhausted T cells (Tex^prog), These are the less differentiated, stem-like exhausted cells. They express the transcription factor TCF1 (also known as TCF7) and other markers like CXCR5, and SLAMF6. They are PD-1 positive but TIM-3 negative (TIM-3 is an exhaustion marker that comes up later). Tex^prog cells can self-renew and have multipotency: they can proliferate and spawn more differentiated progeny. However, on their own they are not highly cytotoxic, think of them as the reservoir or seed population.
Terminally exhausted T cells (Tex^term), These are the end-stage exhausted T cells. They typically have lost TCF1 expression and instead express markers of terminal differentiation like TIM-3, CD39, CD101, and often high levels of inhibitory receptors (CTLA-4, TIGIT, etc.). Tex^term cells may still produce some cytotoxic molecules (in fact, some studies show they can have high levels of granzyme B and even be initially “polyfunctional” in terms of cytokines), but critically, they lack proliferative capacity and durability. They are often short-lived effector cells on their way out. Once a T cell reaches this terminal exhausted state, it can’t revert to a memory state or significantly expand; it’s essentially the end of the road.
Another way to think of it: Tex^prog are akin to stem-like memory cells residing in a tumor, whereas Tex^term are akin to overworked effector cells that have seen too much antigen. The presence of TCF1 (in Tex^prog) is a key indicator because TCF1 is known to be vital for T cell development and memory formation. The Tex^prog vs Tex^term distinction is not just a surface marker curiosity, it comes with profound functional differences and different roles in therapy response, as we’ll explore.
Why “Burned-Out” T Cells Underperform
Relying on terminally exhausted T cells to do the heavy lifting in cancer therapy is a bit like banking on exhausted workers to pull off an all-nighter, it’s usually a losing proposition. By the time CD8⁺ T cells become Tex^term, they have undergone extensive epigenetic reprogramming that locks in their dysfunction. They remain stuck expressing inhibitory checkpoints and are refractory to further activation. In fact, research shows that exhausted CD8 T cells acquire a unique chromatin landscape, and even if you remove the chronic antigen stimulus (e.g. cure the infection or presumably remove tumor), the epigenetic “scars” of exhaustion persist. In a 2021 study of patients who had chronic hepatitis C or HIV infections that were cured, the virus-specific CD8 T cells still retained an “exhausted” epigenetic program and did not revert to fully functional memory cells. In other words, terminal exhaustion is hard-wired in many ways; once that cell is terminally exhausted, it’s not easily resurrected to a youthful state. This explains why simply blocking PD-1 can fail if a T cell is too far gone, the checkpoints are just the tip of the iceberg; the cell’s whole gene expression and epigenetic profile have been altered in exhaustion.
Functionally, terminally exhausted T cells exhibit poor proliferation and survival. They’ve often upregulated the transcription factor TOX and downregulated memory-associated factors like TCF1 and IL7R, cementing a terminal fate. They are prone to apoptosis and lack responsiveness to homeostatic cytokines. For instance, exhausted CD8 T cells have an impaired ability to proliferate even if you give them IL-7 or IL-15 (cytokines that normally help T cells survive and divide). They essentially require the antigen to hang around at all, and even then, the bulk of the exhausted cells won’t multiply.
It’s true that Tex^term cells can show some immediate effector function, e.g. in some tumors, the terminal subset might make IFN-γ or have granzyme to kill a target cell. But this comes at a steep cost: they don’t persist. They “exert a stronger antitumor response” in the very short term but cannot sustain the immune pressure, they flame out quickly and cannot be reinvigorated by therapies. Notably, because they lack TCF1, they do not proliferate in response to PD-1 blockade. Checkpoint inhibitors like anti-PD-1 basically bounce off these cells, there is little “gas left in the tank” to ignite. Indeed, studies in mice have shown that PD-1 therapy primarily works by acting on the progenitor exhausted cells, not the terminal cells; the progenitors proliferate and differentiate into fresh effectors upon PD-1 blockade, whereas the terminal cells contribute minimally.
Another major limitation of Tex^term cells is their susceptibility to the tumor microenvironment’s suppressive signals. By the time a cell is terminally exhausted, it often expresses high levels of multiple inhibitory receptors simultaneously, PD-1, CTLA-4, TIM-3, TIGIT, LAG-3, you name it, making them sitting ducks for suppression by regulatory T cells, macrophages, and tumor cells that produce the corresponding ligands. In fact, a high abundance of Tex^term cells in tumors has been correlated with the presence of more Tregs (immunosuppressive regulatory T cells), suggesting that the same conditions that drive T cells into terminal exhaustion (chronic antigen, TGF-β, IL-10, etc.) often come hand-in-hand with lots of Tregs. Tregs further promote T cell exhaustion by inducing more inhibitory receptor expression on CD8 T cells, creating a vicious cycle. So a tumor full of terminally exhausted CD8s is typically an immunosuppressed one.
Finally, from a therapeutic development standpoint, targeting terminally exhausted T cells is a bit like targeting a symptom rather than the cause. Sure, one could try to develop drugs that “reprogram” these cells (e.g. epigenetic drugs that erase exhaustion marks, or cytokines to transiently boost them), but you’d be fighting an uphill battle against the cell’s biology. As one review succinctly noted: although in its terminal stage, exhaustion is epigenetically fixed and cannot be reversed; the better approach is to prevent T cells from reaching that terminal stage in the first place. This is where progenitor T cells come in, offering a chance to intervene earlier in the differentiation path and maintain the T cell population in a more favorable state.
The Regenerative Force in Tumors
If terminally exhausted T cells are a dead end, progenitor exhausted CD8⁺ T cells are the renewable energy source that we can tap into for sustained immunity. These cells are often identified by the marker TCF1 (T cell factor 1), a transcription factor that is crucial for T cell self-renewal and memory. In tumors and chronic infections, the TCF1⁺ subset of CD8 T cells (Tex^prog) behaves almost like a cadre of stem cells: they self-renew (maintaining their own population) and periodically differentiate into more potent effector T cells (which can immediately attack the tumor). This two-tier system, a self-renewing pool plus a terminally differentiating output, is reminiscent of how bone marrow stem cells continuously supply short-lived blood cells. Here, the TCF1⁺ progenitor T cells continuously supply exhausted effector cells. It’s an ongoing cycle of regeneration that can keep an anti-tumor immune response going, if the progenitor pool itself is preserved.
Key advantages of progenitor-like CD8 T cells:
They can proliferate. Tex^prog cells are not at the end of their replicative life. Give them the right signals (antigen stimulation with co-stimulation, or release their brakes with PD-1 blockade) and they undergo a proliferative burst, generating a wave of new daughter T cells. For example, after anti-PD-1 therapy in mice, it’s been shown that the PD-1⁺ TCF1⁺ cells divided asymmetrically: one daughter became an effector (TCF1⁻) and the other re-expressed TCF1 to remain as a progenitor, thus the pool sustains itself even as it produces fighters. This is a beautiful mechanism: checkpoint blockade doesn’t so much “wake up” the exhausted effectors as it recruits fresh troops from the progenitor pool, all while keeping that pool from running dry. From an investor’s perspective, therapies that expand this population are effectively amplifying the patient’s own T cells in vivo, akin to an on-board T cell factory, instead of relying only on the finite cells present at baseline.
They maintain killing potential over the long term. Although progenitor exhausted T cells are not big killers themselves (they typically have lower immediate cytotoxicity and only modest IFN-γ production individually ), they are the ones that give rise to the killer cells when needed. Think of Tex^prog as generals or strategists, relatively calm in the moment, that can spawn an army of foot soldiers (Tex^term effectors) when the battle heats up. Without those generals, once the initial soldiers tire out, there’s no one to call reinforcements. Indeed, experiments have demonstrated this starkly: if you isolate just the progenitor-like T cells from a tumor and transfer them into a tumor-bearing mouse, they control tumor growth far better than an equal number of terminally exhausted cells. In one study, mice with melanoma received TILs sorted into TCF1⁺ (progenitor) vs TCF1⁻ (terminal) subsets, the mice given the TCF1⁺ subset had substantially reduced tumor growth compared to those given the terminal T cells. The progenitors could continue to proliferate and differentiate in the new host, whereas the terminal cells were largely end-of-line.
They are essential for immunotherapy success. Mounting evidence indicates that the presence of a robust Tex^prog compartment is a prerequisite for effective response to checkpoint inhibitors like PD-1/PD-L1 blockers. For instance, in patients with melanoma, the frequency of TCF1⁺ PD-1⁺ CD8 T cells in the tumor has been associated with better outcomes on PD-1 therapy. One study noted that melanoma tumors from responders had higher fractions of these progenitor-exhausted cells, and these patients enjoyed longer progression-free and overall survival under checkpoint therapy. Similarly, in lung cancer and other tumors, spatially resolved analyses have found that TCF1⁺ stem-like T cells localize in particular niches and correlate with response to treatment. We will discuss the spatial aspect shortly, but the high-level takeaway is: checkpoint blockade doesn’t magically make every exhausted T cell functional, it unleashes the progenitors to refuel the attack. Without progenitors, drugs like nivolumab or pembrolizumab have very limited targets to act on, which might explain non-responders.
Long-term immunity and memory formation. Progenitor CD8 T cells also have the molecular toolkits associated with forming true memory cells. They express genes involved in survival (like IL-7R) and retain the ability to persist long-term. With appropriate CD4⁺ T cell help and dendritic cell signals, some of these TCF1⁺ cells might even become bona fide memory T cells in the tumor that stick around after tumor clearance. This is crucial for preventing relapse, you want a persistent immune surveillance. Patients who have those supportive interactions (often facilitated by CD4 “helper” T cells) develop better CD8⁺ T cell memory and are less likely to relapse. In short, progenitor-like T cells are the bridge between an active fight and immune memory, they are the cells that can remember the tumor in the long run.
Given these advantages, it’s clear why targeting or expanding the progenitor pool is a superior strategy. Rather than pushing already-exhausted cells to do what they can’t, this approach invests in the body’s own capacity to regenerate the immune attack. It’s akin to improving the stem cells in a bone marrow transplant rather than transfusing short-lived cells over and over. Now, how do we do this in practice? That’s where certain therapeutic strategies come in.
Sustaining the Progenitor Pool
Understanding that progenitor T cells are the “golden geese” of anti-tumor immunity, many emerging therapies aim to keep these cells alive, thriving, and differentiating at the right pace. Several approaches have shown promise:
IL-15 Agonists, Fuel for Memory and Stem-Like T Cells: IL-15 is a cytokine known for its role in maintaining memory CD8 T cells in the body. Unlike IL-2 (which can drive T cells to differentiate and can also expand Tregs, with a risk of exhaustion and toxicity), IL-15 mainly supports homeostatic proliferation and survival of T cells and NK cells without Treg expansion. In the context of exhaustion, IL-15 has demonstrated a remarkable effect: it preferentially expands the progenitor-like T cells. A 2023 study showed that ex vivo IL-15 exposure caused a significant proliferation of TCF1⁺ progenitor exhausted CD8 T cells, while the more terminal subsets proliferated much less. In chronically infected mice, giving IL-15 in vivo boosted the self-renewal of the Tex^prog pool in spleen and bone marrow, effectively increasing the reservoir of T cells available to fight. Importantly, this translated to human tumors as well: CD8⁺ tumor-infiltrating lymphocytes from renal cell carcinoma patients, when cultured with IL-15, showed a preferential expansion of the TCF1⁺ subset compared to the terminally exhausted subset.
This understanding has led to IL-15 “superagonists” being tested in cancer. N-803 (Anktiva) is one such IL-15 receptor agonist complex that has gained attention. It effectively mimics IL-15 but with enhanced pharmacokinetics and potency. In April 2024, the FDA approved N-803 in combination with BCG (an immunotherapy for bladder cancer) for certain non-muscle invasive bladder cancers. The rationale is that IL-15 can sustain the body’s T cell and NK cell populations post-BCG, driving durable responses. Early results are encouraging, a significant proportion of patients had long-lasting complete responses, indicating the immune system remained engaged over time. Beyond bladder cancer, IL-15 agonists are being explored as adjuvants to checkpoint inhibitors and cancer vaccines, precisely because they can expand that critical pool of stem-like T cells and potentially even rescue exhausted T cells from oblivion by giving them a proliferative and survival boost.
For investors, it’s noteworthy that IL-15 agonists represent a tissue-agnostic immunotherapy component: they aren’t specific to one cancer antigen or type, but rather enhance the immune milieu universally. That means they could be combined with various therapies (checkpoint blockers, CAR-T, bispecific engagers) across many cancers, a broad market opportunity.
Co-stimulation and the Full Activation Signal: T cells require not just one but three signals to mount a sustained response: Signal 1 is antigen via the TCR, Signal 2 is co-stimulation (e.g. CD28 or 4-1BB provided typically by an antigen-presenting cell), and Signal 3 is cytokine support (like IL-15 or IL-12). One reason T cell engagers (e.g. bispecific antibodies that directly bind a T cell to a tumor cell) have had limited durability in solid tumors is that they often provide only Signal 1 (TCR stimulation via CD3). Without co-stimulatory Signal 2 and cytokine Signal 3, CD8 T cells rapidly exhaust after an initial burst of activity. They kill some tumor cells, then essentially collapse, unable to persist or proliferate further.
To counter this, next-generation engagers and cell therapies are incorporating co-stimulatory signals. For example, some bispecific T cell engagers are being engineered with attached co-stimulatory domains or given in combination with 4-1BB agonists. Co-stimulation helps license the T cell to survive and divide, not just kill. A vivid illustration came from in vivo imaging of immune responses: when CD8 T cells in tumors received help from CD4⁺ T cells and dendritic cells (which together supply CD28 ligation and cytokines), those CD8 T cells avoided exhaustion and kept functioning over time. In immune “triads” observed in tumor tissues, structures where a CD8 T cell, a CD4 helper T cell, and a dendritic cell cluster together, the CD8 T cells were able to differentiate into potent killers and form memory, rather than flaming out. These triads essentially showcase co-stimulation and cytokine support in action, right in the tumor microenvironment. In patients who responded well to therapy, such structured immune cell interactions were common; in non-responders, they were absent.
What this means for therapy design is that drugs which encourage co-stimulation, or at least avoid blocking it, will sustain the T cell response better. Checkpoint inhibitors indirectly do this (by removing inhibitory signals, they let costims work unopposed). Now, companies are looking at agonist antibodies for co-stimulatory receptors (like CD28, 4-1BB, OX40, CD27) as adjuncts to checkpoint blockade. For example, there are trials of anti-PD-1 combined with an agonist anti-CD28, the idea being to push the progenitor T cells into dividing and the effector T cells into killing mode, without inducing exhaustion. It’s a fine balance (too much stimulation can also cause T cell burnout), but done correctly, perhaps in a spatially restricted way within tumors, it could dramatically increase the pool of functional T cells on the ground.
Epitope Spreading and Vaccination Effects: An often under-appreciated way to sustain the anti-tumor T cell pool is by enabling epitope spreading. Epitope spreading is essentially the immune system’s version of adaptative learning, once it starts killing tumor cells, it can pick up new antigen fragments from those cells and generate fresh T cell responses to different tumor antigens. This creates a broader, polyclonal T cell response that the tumor will have a much harder time escaping. But epitope spreading won’t happen if all T cells rush to terminal exhaustion after the first round of killing. To achieve spreading, you need active dendritic cells presenting antigens and you need a supply of T cells capable of responding to those new antigens (which might be the progenitor T cells or new T cells recruited from lymphoid organs).
Therapies that promote epitope spreading include cancer vaccines and oncolytic viruses (which introduce new antigens and inflammation), as well as combination approaches that involve cell death (to release antigens) plus immunostimulation. The earlier-mentioned immune triads are thought to be hubs of epitope spreading, dendritic cells in triads were observed picking up tumor antigen debris and presenting it to T cells, while nearby CD4 helpers boosted the CD8s’ ability to respond. The result was a real-time broadening of the immune response and adaptation to tumor antigen changes. Clinically, this translates to longer remissions and less likelihood of relapse due to antigen escape. From a pan-cancer perspective, enabling T cells to “spread” their targets means a single therapy can effectively sic the immune system on many different mutations or antigens in a tumor, a particularly valuable trait in tumors with high heterogeneity or mutational burden.
Tumor Microenvironment (TME) Modulation: The TME heavily influences whether T cells become exhausted or stay functional. As noted, factors like hypoxia, immunosuppressive cells (Tregs, MDSCs), and certain cytokines (TGF-β, IL-10) all drive T cells toward an exhausted phenotype. Therefore, another strategy is to modulate the TME to be more permissive for progenitor T cells and less conducive to terminal exhaustion. Two examples:
Anti-angiogenic therapy: Tumor hypoxia is a known driver of exhaustion and dysfunction. Hypoxia stabilizes HIF-1α, which in T cells upregulates inhibitory checkpoints and blunts effector functions, while also causing tumor cells to secrete VEGF-A, a factor that not only creates aberrant vasculature but also directly promotes CD8 T cell exhaustion via TOX upregulation. Anti-angiogenic drugs (like bevacizumab or TKIs against VEGF receptors) can normalize blood flow and reduce hypoxia. By doing so, they may indirectly reduce that exhaustion pressure. In fact, research indicates VEGF-A blockade can prevent some of the terminal differentiation of T cells and improve their response to immunotherapy. Clinically, combinations of anti-VEGF agents with PD-1/L1 blockade (e.g. bevacizumab + atezolizumab in renal cell carcinoma, or lenvatinib + pembrolizumab in endometrial carcinoma) have shown improved outcomes, likely partly because the TME is made more supportive for T cells. Essentially, a less hypoxic, less VEGF-rich tumor is one where progenitor T cells might survive longer and not be pushed to exhaustion so quickly.
Targeting immunosuppressive cells: As mentioned, Tregs foster exhaustion. Depleting or functionally blocking Tregs can free up CD8 T cells to remain active. This is one reason anti-CTLA-4 (ipilimumab) can synergize with PD-1 inhibitors, beyond its checkpoint function, CTLA-4 blockade preferentially depletes Tregs in the tumor (through ADCC mechanisms), relieving a source of immunosuppression. The net effect is a more favorable environment for CD8 T cells to maintain their proliferative, cytokine-producing capacity. Other approaches include targeting adenosine pathways (as adenosine in TME can put T cells to sleep), or CSF1R blockers to reduce suppressive tumor macrophages. All these are TME mods that aim to prevent the premature terminal differentiation of T cells. For example, if you reduce TGF-β in a tumor (perhaps via small molecule or antibody), T cells don’t upregulate as many inhibitory receptors and can retain a more functional, less exhausted phenotype.
Building supportive niches: The flip side of removing bad factors is adding good ones. Some therapies aim to increase the presence of dendritic cells or local lymphoid structures in tumors. One fascinating observation from spatial immunology is that TCF1⁺ progenitor T cells often reside in or near tertiary lymphoid structures (TLS) within tumors. TLS are like lymph node–like patches that form in some tumors, they contain B cells, T cells, dendritic cells, and even blood vessels, creating a mini-immune hub on the tumor site. These structures appear to be safe havens for progenitor T cells: a recent high-plex imaging study in esophageal cancer found that TCF1⁺ CD8 T cells (especially those co-expressing an activation marker CD39) were concentrated within and around TLS, predominantly in the stromal areas of the tumor. Tumors that had TLS contained significantly more of these progenitor T cells than tumors that lacked TLS. This suggests the TLS provide the nurturing environment, likely via dendritic cell interactions and IL-7/IL-15, to sustain the progenitor pool. Moreover, patients with TLS-rich tumors often respond better to immunotherapy. In the esophageal study, those distinct CD39⁺ TCF1⁺ T cells correlated with benefit from PD-1 blockers. Another study in lung cancer noted that PD-1⁺ TCF1⁺ stem-like CD8 cells were preferentially found in TLS regions of the tumor, reinforcing that observation.
Therefore, therapies that can induce or expand TLS in tumors (e.g. certain intratumoral injections, TLR agonists, or even local radiotherapy which can cause an inflammatory lymphoid aggregation) might indirectly increase progenitor T cell numbers. It’s a cutting-edge area of spatial immunology: essentially landscaping the tumor microenvironment to include “homes” for T cells. If successful, this could turn previously cold, inhospitable tumors into environments where an army of T cells can both generate (in the TLS) and then go out to fight in the tumor parenchyma.
In summary, sustaining the progenitor pool is about providing the right growth conditions for T cells: the cytokines (like IL-15), the co-signals (CD28/ICOS, etc.), the supportive cell types (CD4 T helper, cDC1 dendritic cells), and the physical niches (TLS). All these ensure that the anti-tumor immune response isn’t a brief blaze, but a long-burning fire.
Smarter Immunotherapy for Solid Tumors
Why is this especially exciting from a pan-cancer investment perspective? Because the principles of T cell differentiation apply across essentially all solid tumors. Nearly every type of cancer where T cells matter, melanoma, lung cancer, renal cell carcinoma, head and neck cancer, liver cancer, etc., shows this dichotomy between progenitor and terminal exhausted T cells in the tumor. Thus, a therapy that can tilt the balance toward progenitors or improve their survival and function has the potential for broad use. We’re talking about mechanisms, not a single antigen target, meaning such an approach can layer on top of many modalities.
Let’s consider a few cornerstone immunotherapy modalities and how focusing on progenitor CD8 T cells enhances them:
Checkpoint Inhibitors: These drugs (PD-1/PD-L1, CTLA-4 blockers) have given us a first taste of durable cancer control in a subset of patients. We now understand that responders often have pre-existing T cell populations capable of being revitalized, specifically, those TCF1⁺ progenitor cells. To broaden checkpoint therapy’s success, the goal is either to increase those cells in patients who lack them (perhaps by a priming vaccine or IL-15, etc.) or to maintain them during therapy. This could reduce cases of relapse: for instance, some patients initially respond to PD-1 but then progress, possibly because the progenitor pool was small and got exhausted or depleted after initial response. If we co-treat with something that keeps that pool thriving, we might see more sustained remissions. There is also interest in using biomarkers like TCF1⁺ CD8 counts in tumors to select patients or track responses. In the coming years, when evaluating an immunotherapy investment, one might ask: does this approach preserve the TCF1⁺ population or drive terminal differentiation? The ones that preserve the “generative capacity” of the T cell response will likely produce longer-lasting clinical benefits.
T Cell Engagers and CAR-T for Solid Tumors: CAR-T cell therapy has worked wonders in blood cancers but not yet in solid tumors. One reason: CAR-T cells often terminally differentiate and exhaust in the harsh solid tumor microenvironment. Engineering CAR-T or TCR-T cells to have a more memory or progenitor phenotype is a hot area of research. Some groups are tweaking CAR co-stimulatory domains to favor memory formation (e.g. using 4-1BB costim yields more memory-like CAR T cells than CD28 costim, which tends to drive more effector differentiation). Others add small molecules or cytokines during CAR-T expansion to enrich for TCF1⁺ stem cell memory T cells. The idea is to infuse a product that behaves more like a self-renewing population than a bolus of effector cells. Likewise, for bispecific T cell engagers, companies are now exploring dual-target molecules that engage CD3 and provide a co-stimulatory signal (like binding CD28 on T cells at the same time). These could prevent the rapid exhaustion seen with first-generation T cell engagers. As noted earlier, the next wave of engagers will likely incorporate these lessons, ensuring the T cells they activate don’t “lose steam quickly” in solid tumors. For investors, any platform that can show T cells remaining functional and proliferative in the tumor over time (perhaps via serial biopsies or advanced imaging) will be very compelling.
Cancer Vaccines and Adoptive T Cell Transfer: Even cancer vaccines (which aim to generate new T cell responses from scratch) should ideally produce T cells that enter the tumor as Tex^prog rather than immediately becoming Tex^term. Vaccine regimens might be paired with immune modulators to encourage that outcome. Similarly, when we transfer tumor-infiltrating lymphocytes (TIL) or neoantigen-specific T cells into patients, we might consider pre-selecting or enriching for the progenitor phenotype cells in the lab (some evidence suggests the TIL subsets with high stem-like qualities engraft and mediate tumor regression better). This could improve the durability of cell therapy responses, making TIL therapy more consistently effective across cancers.
Spatial Biomarkers and Combination Strategies: High-plex spatial imaging (e.g. imaging mass cytometry, multiplex immunofluorescence) is emerging as a tool to identify whether a patient’s tumor has the right immune “architecture” for success. We discussed TLS and immune cell clustering. If a tumor has none of these structures, perhaps combining a checkpoint inhibitor with an agent known to induce lymphoid chemokines (like LIGHT or CXCL13) could be an approach, literally trying to encourage formation of immune niches. There’s an interplay here: sometimes the therapy itself can create a more favorable microenvironment (e.g. local radiation can cause inflammation that attracts dendritic cells and T cells, potentially seeding TLS; certain chemotherapies at low dose can modulate Tregs or MDSCs). A pan-cancer strategy might involve tailoring combinations to ensure that the progenitor T cells have the support they need in each tumor type. For instance, in a tumor known to be highly VEGF-rich (like renal cancer), adding anti-VEGF to immunotherapy may be critical ; in a tumor with heavy Treg infiltration (like gastric cancer), adding a Treg-targeting agent might be key.
Reduced Relapse and Adaptive Immunity: Ultimately, focusing on progenitor T cells aligns with the goal of turning cancer into a manageable or even curable condition. If therapy leaves behind a vigilant immune memory, the chances of the cancer returning drop dramatically. This is the holy grail: a patient’s immune system that not only eradicates the tumor but “remembers” it and suppresses any resurgence or new lesions (akin to how the immune system handles many viruses post-vaccination or infection). Progenitor T cells, given their capacity for long-term persistence and replenishment of effectors, are the cornerstone of this vision. We are essentially trying to induce a state of tumor-specific immune memory across the body. Achieving that would mean pan-cancer immunotherapy regimens that potentially cure patients or at least give decade-long disease-free intervals, outcomes that investors and companies dream about.
In summary, the shift toward targeting progenitor CD8⁺ T cells represents a maturation of the immunotherapy field’s thinking. Early on, the focus was “more killing, more activation at all costs”, which led to short-lived expansions and also severe toxicities in some cases. Now, it’s “smart activation, sustained coordination”. By engaging the body’s renewable immune capacity, we stand to gain efficacy that not only cuts across tumor types (because T cells are universal fighters) but also lasts far longer, potentially eliminating tumors with lower chances of relapse.
A New Paradigm, Invest in the Immune “Stem Cells,” Not the Exhausted Soldiers
Immuno-oncology is increasingly recognizing that quality trumps quantity when it comes to T cells. Terminally exhausted T cells, with all their inhibitory baggage and irreparable fatigue, are not the ideal assets to bet on. The superior therapeutic opportunity lies in targeting and expanding the progenitor-like CD8⁺ T cells, the cells that can both attack now and regenerate for later. This approach aligns with everything we know about achieving durable immunity: it’s about creating a self-sustaining, adaptable immune response rather than a one-time attack.
For biotech investors and startup founders, this has tangible implications in pipeline decisions and trial designs:
Treatments that preserve TCF1⁺ CD8 T cells or increase their numbers in tumors could become the backbone of combination regimens. We’re already seeing this with IL-15 superagonists and will likely see more such “immune rejuvenating” agents.
Spatial and phenotypic biomarkers like progenitor T cell density or TLS presence will guide patient selection, making trials more efficient and increasing success rates by matching the right strategy to the right immune context.
The concept is inherently pan-cancer, it’s about empowering the patient’s immune system, which is a universal approach not limited by tumor origin. This means potential label expansions and broader markets if a drug proves its worth in one indication.
It also dovetails with the trend of tissue-agnostic approvals: just as we saw checkpoint inhibitors gain tumor-agnostic approval for MSI-high or high mutational burden cancers, a therapy that boosts progenitor T cells might find use across multiple cancers with a common immune phenotype.
In the competitive landscape, those who incorporate this T cell differentiation insight will design smarter drugs. It could be the differentiator between an immunotherapy that produces a few months of tumor control versus one that produces years of remission. As one thought leader encapsulated, the future of T cell therapy isn’t about hitting harder, it’s about engaging smarter. Engaging smarter means engaging the full immune circuit, activating T cells with their natural allies (helper T cells, dendritic cells) and not pushing them into oblivion but keeping them in the game for the long run.
By targeting progenitor CD8⁺ T cells and the factors that nurture them, we align our therapies with the immune system’s own design for durable defense. It’s a shift from a sprint to a marathon mindset, and for patients, it could mean the difference between a transient response and a cure. For investors, backing this approach means investing in treatments with the potential for deep, lasting efficacy across a broad swath of cancers, and that is indeed an immunotherapy goldmine in the making.