Trastuzumab Resistance in HER2+ Breast Cancer: When Cancer Rewires the Network

Why resistance can emerge through interconnected changes in receptors, signalling, transcription, metabolism and the tumour microenvironment.

Trastuzumab transformed the treatment of HER2-positive breast cancer.

By targeting the HER2 receptor, it provided one of the clearest demonstrations of how understanding a molecular driver of cancer can lead to an effective targeted therapy. Yet HER2-positive tumours are not static systems. Some respond poorly from the outset, while others initially respond and later develop resistance.

Increasingly, research suggests that there may not be one answer. Trastuzumab resistance can emerge through several interconnected molecular routes, making it better understood as a reorganisation of cellular regulatory networks rather than simply a change in one gene or pathway.

In this article

HER2 is the target — but the biology extends far beyond HER2

HER2 belongs to the ERBB family of receptor tyrosine kinases. When HER2 signalling is strongly activated, downstream pathways controlling proliferation, survival, metabolism and cell-state regulation can become abnormally active.

Trastuzumab binds to the extracellular domain of HER2. Its effects include interfering with HER2-associated signalling and engaging immune mechanisms such as antibody-dependent cellular cytotoxicity.

But cancer cells have multiple opportunities to circumvent this pressure. Some resistance mechanisms alter the target itself. Others reactivate pathways downstream of HER2, activate alternative receptors, remodel transcriptional programmes or change the metabolic and immune environment surrounding the tumour.

This diversity is precisely what makes resistance difficult to understand from a single molecular measurement.

1The target itself can change

One route to resistance involves changes affecting HER2 recognition or signalling.

For example, truncated forms such as p95HER2 can lack portions of the extracellular domain targeted by trastuzumab while retaining signalling activity. Other mechanisms can physically reduce antibody access to HER2; proteins including MUC4 have been implicated in masking the receptor.

HER2 mutations and alternative receptor forms may also alter signalling behaviour and therapeutic sensitivity.

In these situations, the tumour may remain biologically dependent on HER-family signalling while becoming less vulnerable to the original therapeutic intervention.

2Cancer cells can find alternative signalling routes

Even when HER2 remains effectively targeted, downstream signalling does not necessarily stop.

The PI3K-AKT-mTOR pathway is one of the central survival pathways associated with HER2-positive breast cancer. Alterations affecting components such as PI3K or PTEN can allow downstream signalling to remain active despite inhibition at the receptor level.

Other receptors can also provide bypass routes. IGF-1R, HER3, EGFR, MET and AXL have all been investigated as potential contributors to alternative signalling in resistant cells.

The important concept is that a signalling network has redundancy. Blocking one entrance does not necessarily shut down the system if the cancer cell can activate another route to the same downstream survival machinery.

3Resistance can involve a change in the transcriptional programme

Signalling pathways ultimately influence gene expression.

This means that resistance may become established not simply because one signalling molecule becomes activated, but because the cell adopts a different regulatory state.

Transcription factors can coordinate hundreds of genes associated with proliferation, stress responses, inflammation, differentiation and metabolism. Changes in chromatin accessibility can further alter which regulatory regions are available for transcription-factor binding.

Once such a programme becomes established, the resistant cell may behave very differently from the sensitive cell from which it originated.

This is one reason why a list of differentially expressed genes, although extremely useful, may not fully explain resistance. The next question is: Which regulatory mechanisms are producing those expression changes?

4Metabolism can become part of the escape strategy

Therapeutic resistance also has a metabolic dimension.

Cancer cells continuously adjust how they generate energy, synthesize cellular components and respond to stress. HER2-positive cells that escape treatment can exhibit altered glycolysis, mitochondrial activity, lipid metabolism and metabolic-signalling interactions.

Recent work increasingly describes metabolic reprogramming as part of the adaptive architecture of trastuzumab resistance rather than simply a secondary consequence of tumour growth.

This is important because signalling, transcription and metabolism do not operate independently. Each can reinforce the others. A signalling alteration may change transcription. A transcriptional programme can reprogramme metabolism. Metabolic changes can then support survival and feed signals back into the regulatory network.

5The tumour microenvironment matters too

Trastuzumab does not act solely through intracellular signalling inhibition. Immune-mediated mechanisms contribute to its activity.

Changes in the tumour microenvironment can therefore affect treatment response. Immunosuppressive macrophages, altered NK-cell activity, immune-checkpoint signalling and other microenvironmental factors can reduce effective anti-tumour immune responses.

Current research into HER2-targeted therapy resistance increasingly considers this interaction between the cancer cell and its surrounding microenvironment.

The resistant phenotype may therefore reflect several layers simultaneously:

That makes resistance fundamentally a systems-biology problem.

From “what changed?” to “what is driving the change?”

A particularly useful distinction in resistance research is the difference between identifying altered molecules and identifying regulatory drivers.

RNA-seq might reveal thousands of genes whose expression differs between sensitive and resistant cells. ATAC-seq can reveal changes in chromatin accessibility. Both provide valuable information.

But a mechanistic interpretation requires another level of questioning:

• Which transcription factors could coordinate these genes?
• Which signalling molecules control those transcription factors?
• Are there upstream regulatory nodes that influence several branches of the resistance programme at once?
• Are feedback mechanisms helping maintain the resistant state?

Answering these questions can turn a molecular signature into a testable model of resistance.

A real-data example: tracing the regulatory architecture of trastuzumab resistance

To explore this question, a geneXplain analysis used publicly available multi-omics data from a study by Mukund and colleagues comparing trastuzumab-sensitive BT474 HER2+ breast cancer cells with a resistant derivative, BT474R. The original study combined transcriptomic, chromatin, signalling and metabolic measurements to investigate acquired resistance.

Rather than stopping at differential expression, the geneXplain analysis integrated RNA-seq and ATAC-seq information and worked upstream from altered genes toward their regulatory architecture.

At baseline, the resistant-versus-sensitive comparison identified 2,091 significantly upregulated and 1,998 significantly downregulated genes under the thresholds used in the report. IGF2 was among the strongest expression changes.

Promoter/enhancer and transcription-factor analysis then highlighted regulatory factors including RELA, FOS, RUNX3, RXRA, SOX9 and PARP1. From these transcription factors, upstream network analysis identified candidate master regulators including GDF15/MIC-1, IGF2 and LRRK2, among others.

Notably, IGF2 emerges here from a separate analytical step than the earlier differential-expression result. It is independently flagged as a network node upstream of the identified transcription factors, not simply restated from the expression data.

Importantly, these computational results should be viewed as mechanistic hypotheses and candidates for further experimental investigation, rather than proof that any one molecule alone causes clinical trastuzumab resistance.

What makes the analysis interesting is the strategy: instead of asking only which genes differ, it asks how those genes may be connected through transcription factors, upstream signalling molecules and feedback relationships.

Visual walkthrough

Resistance may be a state, not a single event

A tumour cell can modify receptor behaviour, recruit alternative pathways, change its transcriptional circuitry, alter metabolism and establish feedback loops that reinforce the new state.

This has an important implication for research.

Finding another differentially expressed gene is useful. Finding the regulatory network that explains why hundreds of genes changed together may bring us closer to understanding how resistance develops, how it is maintained and where it might be disrupted.

That shift from molecular lists toward mechanisms is likely to become increasingly important as transcriptomic, epigenomic, proteomic and other omics datasets are studied together.

References and further reading

1. Mukund K, Alva-Ornelas JA, Maddox AL, Murali D, Veraksa D, Saftics A, et al. Molecular atlas of HER2+ breast cancer cells treated with endogenous ligands: temporal insights into mechanisms of trastuzumab resistance. Cancers (Basel). 2024;16(3):553. doi: 10.3390/cancers16030553. PMID: 38339304.
2. Wang L, Wang Y, Li Y, Zhou L, Du J, Wang J, et al. Resistance mechanisms and prospects of trastuzumab. Frontiers in Oncology. 2024;14:1389390. doi: 10.3389/fonc.2024.1389390. PMID: 39655080.
3. Pan L, Li J, Xu Q, Gao Z, Yang M, Wu X, Li X. HER2/PI3K/AKT pathway in HER2-positive breast cancer: a review. Medicine (Baltimore). 2024;103(24):e38508. doi: 10.1097/MD.0000000000038508. PMID: 38875362.
4. Martin-Castillo B, Verdura S, Llop-Hernández À, Lupu R, Cuyàs E, Menendez JA. Metabolic hallmarks of trastuzumab resistance. Expert Opinion on Therapeutic Targets. 2025;29(7):457–479. doi: 10.1080/14728222.2025.2532394. PMID: 40642957.