New Molecular Drivers of HER2 Breast Cancer Resistance: A Q&A With Ateequllah Hayat, PhD

Approximately 70% of patients with breast cancer relapse after 5 years of treatment, thus incentivizing new research that addresses drug resistance and how it translates into poor prognosis, specifically in patients with HER2-positive breast cancer.¹

Researchers at City St. George’s, University of London, have identified a new molecular signature that may explain why patients with aggressive HER2-positive breast cancer developed resistance to drugs like lapatinib—a dual kinase inhibitor targeting HER2 and epidermal growth factor receptor (EGFR).

The corresponding author of a new study, Ateequllah Hayat, PhD, lecturer in drug development at City St. George’s, spoke with The American Journal of Managed Care^® (AJMC^®) to discuss his team’s findings and future applications for this research.

This transcript has been lightly edited for clarity.

AJMC: What are the key hidden drivers of drug resistance in aggressive breast cancer that your team identified, especially involving epigenetics and noncoding RNAs?

Hayat: We took some breast cancer cells from HER2-positive breast cancers, and we made them resistant to a targeted therapy called lapatinib, and then we were able to do a multiomics approach, essentially looking at epigenetic changes, transcriptomic changes, and proteomic changes. What we found was quite interesting in this model: you see the opening up of the DNA chromatin that is associated with oncogenic activation, but rather than the classical opening up associated with oncogenic activation, we instead saw that lapatinib-resistant cells displayed globally reduced chromatin accessibility. That means that the DNA within these lapatinib-resistant cells actually became more compacted.

However, when we had a high-resolution look at some key, specific genes, we saw there were focal increases near specific genes that drive resistance. We think this suggests an epigenetic remodeling process that selectively activates resistance programs while suppressing other regulatory networks. In terms of our gene expression changes, we performed RNA sequencing, and we found there were at least 9 gene signatures that included some of these biomarkers that were already known—for example, EGFR or SCIN—but there were others that were not previously implicated in this HER2 drug-resistance context.

An example of this would be HPGT or FASN, which are both essentially genes or proteins. A lot of these genes were linked to stress responses or remodeling of actin and metabolic rewiring, which are key adaptive phenotypes under therapeutic pressure. The other thing was the epigenetic modulator shift. Differential expression of a number of histone-modifying enzymes, including enzymes called histone deacetylases [HDACs] or DNA methylases, was observed, highlighting that the broader epigenetic regulators are altered during resistance acquisition.

AJMC: You also mapped major epigenetic and transcriptomic changes in lapatinib-resistant tumors. Which of these alterations do you see as the most promising targets for future therapies?

Hayat: We identified a number of epigenetic regulators. As we said, HDACs and other epigenetic enzymes suggest there are epigenetic modulators or HDAC inhibitors that could resensitize these resistant cells to therapies, and that would be the ultimate goal. We also looked at and identified some metabolic enzymes, like the FASN protein that I mentioned briefly earlier, which is consistently upregulated across all 3 state-of-the-art technologies that we used. It also has a pharmacological inhibitor that is in development, and perhaps it is positioning it as a druggable metabolic vulnerability.

Some of the other key pathways that became activated in this, in these screenings or in our analysis with the stress response and cytoskeletal regulators, for example, genes like HPGD or PCP4. It points or indicates toward adaptive survival and motility pathways that facilitate resistance and invasion, potentially targeting it via modulating stress signaling or motility networks.

We also identified previously known genes or receptors. One of them is EGFR, which reaffirms that bypass signaling pathway. Sometimes, when drugs become resistant to HER2, you get this navigation or bypassing of the signaling, not through the HER2 pathway but through its family members like EGFR, which can sustain cell growth despite HER2 being blocked, which supports combination targeting of multiple receptor tyrosine kinases concurrently.

AJMC: Given these hidden drivers, what diagnostic tools (eg, liquid biopsy, methylation panels) could be developed to identify resistance early in the treatment course?

Hayat: If we validate this signature in other cancer models and in patient samples, and if we are able to knock out some of those genes and really get to grips with the fact that this is actually really, truly involved in drug resistance, we could develop liquid biopsy-based assets. Detecting these changes in expression or epigenetically linked signatures from circulating tumor DNA or RNA could signal emerging resistance before clinical progression.

The other thing would be to develop methylation or epigenetic-based panels. These would be targeted panels looking at differential methylation or differentially expressed genes at key regions within the DNA that is implicated in resistance, which could provide early warning of the fact that there's epigenetic rewiring happening toward resistance.

AJMC: How might combination therapies (eg, epigenetic modulators with HER2‑targeted agents) be designed to overcome the resistance pathways you’ve uncovered?

Hayat: The data suggest that the rational combinations target both the epigenetic state and the signaling drivers. The first thing we could test is the epigenetic and the HER2 or the anti-HER2 therapies. Pairing an epigenetic therapy, such as an HDAC inhibitor, with lapatinib could reverse the epigenetic-mediated resistance programs and restore the pattern of sensitivity.

The other thing would be to have dual receptor targeting. I already mentioned that combining HER2 inhibition with its family members, like EGFR or other receptor tyrosine kinase inhibitors, can block those compensatory survival pathways that become activated in drug resistance. Metabolic targeting, such as combining these FASN inhibitors, because FASN came up really highly expressed across 3 of our data sets, with HER2-targeted therapy, might also undermine the metabolic adaptation that sustained these resistant cells.

AJMC: Looking ahead, how do you envision translating these findings into clinical practice, and what hurdles (regulatory, technical, or clinical) do you anticipate?

Hayat: We expect significant hurdles because, again, this is a discovery stage, and there are quite a few stages left to actually get to that. The translation roadmap would essentially look like this. We would have to have diagnostic tools developed, again, bringing those robust assets, whether that's liquid biopsies or tissue-based panels, to clinical labs.

Then we would want to do this clinical validation, which would be to integrate signatures into prospective trials to demonstrate improved outcomes and cost benefits. The therapeutic combination would be the third option, which would be around defining optimal treatments with minimal or manageable toxicities. Regulation is a common hurdle, of course, for many drug development approaches, which will be companion diagnostics that are tied to therapeutic decisions. It would require a high level of evidence, and it would require a significant amount of time and effort to get to that.

Technical issues are the other; standardizing these multiomics assays across the lab in sample types is challenging and very expensive and requires a lot of technicalities in terms of its analysis, as well as using bioinformatics approaches.

Reference

1. Steggall J, Rajeeve V, Al-Subaie N, Naeem A, Ikram A, Naeem A, Hayat A. Integrative proteo-genomic profiling uncovers key biomarkers of lapatinib resistance in HER2-positive breast cancer. BR J Cancer. 2025;133(10):1471-1482. doi:10.1038/s41416-025-03174-3