Beyond “boosting immunity”: cancer stem cells and immunotherapy resistance

Immunotherapy resistance is more than a weak immune response. Discover how changing cancer-cell states, protective surroundings and immune escape connect—and what that means for treatment research.

Page 9 of the hub that unpacks the 2026 Cancer Stem Cells and Drug Resistance Review.

Immunotherapy has changed outcomes for some people with cancer in ways that were once hard to imagine. Immune-checkpoint drugs, cellular therapies and antibody-based treatments can sometimes produce deep, durable responses—even after other treatments have stopped working.

Yet not every cancer responds, and some cancers that initially respond later progress. The 2026 review that initially inspired this study support hub identifies stem-like tumour states, cancer-cell plasticity and the tumour microenvironment as possible contributors to that challenge.

For someone living with progression, this does not mean immunotherapy “does not work,” or that a lack of response means the immune system has failed. It means that successful immune attack depends on a chain of events: the immune system must recognise the cancer, enter the tumour, remain active in the hostile tumour environment and kill enough cancer cells—including cells able to regrow the tumour afterwards. CSC-like cells may interfere with several links in that chain.

A useful image: Immunotherapy is not simply a stronger army arriving at the tumour gates. The immune cells must first find the correct address, recognise the cancer as dangerous, get through the surrounding barriers, avoid being disarmed, and reach the cells most capable of rebuilding the tumour. Stem-like cancer-cell states and the tumour microenvironment may make each of those steps more difficult.

How CSC-like states may evade immunity

The review describes several overlapping ways in which stem-like cancer cells may become harder for the immune system to detect or remove. These mechanisms are not established in every cancer or every patient. They are research frameworks that help explain why immune resistance can be complex.

Possible mechanism	What it means in plain language	Why it may matter
Reduced antigen presentation	A cancer cell may display fewer identifying fragments—antigens—on its surface through MHC/HLA molecules	T cells may have less information to recognise the cell as abnormal
Checkpoint signalling	The tumour can increase “brake” signals such as PD-L1, and may use other immune-suppressive pathways	Even when T cells reach the cancer, they may be switched off or exhausted
Immune-suppressive neighbours	The tumour can recruit or reshape macrophages, Tregs and MDSCs	These cells can reduce effective immune attack and support a protected local niche
Physical barriers	Dense stroma, abnormal blood vessels and extracellular matrix can make entry difficult	T cells and medicines may not penetrate equally into all tumour areas
Metabolic competition	Tumour cells can consume nutrients, create low oxygen and release acidic lactate	The local environment can leave T cells less able to function
Plasticity under treatment pressure	Some cancer cells may shift into a more stem-like, immune-evasive state	Treatment may select for or induce cells that are less visible and harder to kill

These mechanisms overlap. A tumour may simultaneously reduce immune recognition, build a physical barrier, create a low-oxygen and nutrient-poor environment, recruit suppressive immune cells and shift some cells into a treatment-tolerant state. That is why the review argues that targeting the CSC-like cells alone may not be enough; researchers are increasingly considering approaches that also change the immune environment surrounding them.

CSC mimicry: a current idea

One of the newer concepts discussed in this field is CSC mimicry.

This does not mean cancer stem cells are literally pretending to be other cells in a conscious sense. It means cancer cells can activate programs that resemble features of normal stem cells, immune cells, blood-vessel cells or other supportive cell types.

The 2024 Cell Stem Cell review by Saw and colleagues proposes that this flexible “mimicry” may help cancer cells avoid immune detection and survive treatment. For example, stem-like states have been associated in research models with lower MHC-I/HLA-I antigen presentation, altered response to interferon signalling, higher immune-checkpoint signalling and communication that encourages nearby immune or stromal cells to become more tumour-supportive.

In plain language: a cancer cell may change its outward signals, become less recognisable, hide among protective neighbours or create conditions that make immune attack less effective. This is an important hypothesis-driven area of research—not a routine clinical diagnosis and not something that can currently be measured with one standard “CSC mimicry test.”

What this can mean after progression

When immunotherapy does not work, or stops working, the reason may not be one simple “immune escape” switch. It can involve:

The cancer never had features that made it likely to respond.
The immune system could not adequately enter the tumour.
Cancer cells reduced the antigens or MHC/HLA signals that allow T cells to recognise them.
Other tumour cells, immune cells or fibroblasts created an immune-suppressive niche.
Treatment selected for a resistant subclone or a more stem-like, adaptable cell state.
The targeted antigen changed or was lost—particularly relevant to CAR-T and other antigen-directed therapies.[ppl-ai-file-upload.s3.amazonaws][link.springer][karger]

This is why re-evaluation after progression may sometimes be useful. Depending on cancer type, previous testing and treatment history, an oncology team may consider repeat tissue testing, liquid biopsy, biomarker review or clinical-trial assessment. That does not guarantee a new option, and not every person needs further biopsy. But it is reasonable to ask:

“Has anything changed in my tumour biology that could affect immunotherapy, targeted treatment or clinical-trial options now?”

Why standard biomarkers still matter

CSC biology adds nuance; it does not replace evidence-based biomarker testing.

For many cancers, the most actionable current questions may include whether the tumour has:

High microsatellite instability or mismatch-repair deficiency (MSI-H/dMMR).
High tumour mutational burden (TMB-high) in settings where this is relevant.
PD-L1 expression, where it is validated for that cancer and treatment.
A specific genomic alteration that guides targeted therapy or a trial.
Viral, molecular or immune features relevant to the cancer type.

These tests do not perfectly predict benefit. PD-L1 can vary within a tumour and over time; some people with low or absent PD-L1 still benefit in particular cancer settings, while some people with high expression do not. But these biomarkers have far more immediate clinical relevance than measuring a single CSC marker or assigning a generic “immune resistance” score.

Practical question: “Which biomarkers are validated for immunotherapy decision-making in my cancer, have mine been tested, and should any be repeated or expanded after progression?”

What researchers are trying next

The current goal is not simply “find a CSC surface marker and destroy every cell carrying it.” Researchers know that CSC markers can be shared with normal stem cells, can vary between tumour regions and may change as cells change state. A therapy aimed at one marker can also create selection pressure for cells without that marker.

Instead, research is exploring combinations designed to address more than one barrier at once.

Research direction	What it aims to do	Current position
Checkpoint blockade plus CSC/niche targeting	Release immune brakes while disrupting signals that protect CSC-like cells	Highly active research; benefit depends strongly on cancer type and combination
CAR-T and CAR-NK therapies	Engineer immune cells to recognise tumour-associated antigens, including CSC-associated targets	Established in several blood cancers; most CSC-directed and solid-tumour approaches remain early clinical research
Bispecific antibodies	Bring a patient’s own T cells into close contact with tumour cells	Established for some blood cancers; many solid-tumour and CSC-marker approaches are investigational
Multi-target cellular therapies	Recognise more than one antigen to reduce antigen-loss escape	Promising research strategy; safety, persistence and tumour penetration remain challenges
Cancer vaccines	Train immune cells to recognise tumour-associated antigens	A range of approaches are being tested; benefit is cancer- and vaccine-specific
Microenvironment reprogramming	Reduce Treg/MDSC/macrophage suppression, alter fibroblast/stromal barriers or improve immune-cell entry	Strong rationale; most approaches require carefully selected combinations
Metabolic and hypoxia strategies	Make the tumour environment less hostile to immune cells	Mostly early research; the same metabolic pathways also affect normal immune cells

Some of these technologies are already standard treatment in selected blood cancers. For example, CAR-T and bispecific antibodies have transformed options for particular lymphomas, leukaemias and multiple myeloma. But that should not be confused with proof that a CAR-T product targeting a CSC marker will work in solid tumours. Solid tumours pose additional challenges: antigen diversity, poor immune-cell entry, an immune-suppressive microenvironment and potential damage to normal tissue sharing the target.[karger][ascopubs][nature]

A realistic hopeful message

The hopeful point is not that there is a hidden “anti-CSC immunotherapy” ready for everyone. It is that researchers now have a much more sophisticated understanding of why immune responses succeed in some tumours and fail in others.

The field is moving from a question like:

“Can we activate the immune system?”

toward more precise questions:

Which cells is the immune system failing to see?
What barrier prevents immune cells from reaching them?
Which suppressive signals are active in this tumour?
Is there a targetable mutation, antigen or immune feature?
Would a combination strategy make biological sense for this cancer?
Is there a trial that matches the cancer’s current biology—not only its original diagnosis?

For people navigating progression, this supports a grounded form of hope: there may be more to assess than a scan alone. Molecular profiling, biomarker review, pathology review and clinical-trial matching can sometimes identify a different route forward. The right route is not always an immunotherapy, and it will not be available for everyone. But CSC and immune-resistance research is helping create the next generation of more selective, combination-based cancer treatments.[link.springer]

Beyond “boosting immunity”: cancer stem cells and immunotherapy resistance

How CSC-like states may evade immunity

CSC mimicry: a current idea

What this can mean after progression

Why standard biomarkers still matter

What researchers are trying next

A realistic hopeful message

Further reading

Core anchor paper

CSCs and immune escape

EMT, plasticity and immunity

CAR-T and bispecific strategies

On this page