Scoring HER2 by Algorithm: Computational Analysis and the Ultralow Problem

There's a particular kind of humility that sets in when you realize the machine is stuck in exactly the same spot you are. That's roughly where computational HER2 image analysis finds itself in 2025, impressively good at counting membrane staining, and still stumped by the same faint, ambiguous blush of chromogen that divides pathologists reading the same slide.

The development

The immediate trigger for this reckoning is a body of work asking a pointed question: can artificial intelligence make the HER2-low and HER2-ultralow call reproducible where human eyes cannot?

The strongest quantitative signal at the low boundary itself comes from reader studies of AI-assisted scoring: AI assistance raised pathologist agreement on the IHC 0-versus-1+ call from 69.8% to 87.4%, with a standalone algorithm reaching 92.1% accuracy [4], real gains in measurement consistency that nonetheless leave the boundary contested. A June 2025 observational study put the tools to a further test, examining 47 advanced breast carcinoma cases across matched core biopsies, surgical resections, and distant metastases, scoring HER2 status by three methods in parallel, conventional light microscopy, digital pathology on whole-slide images, and AI-assisted image analysis [1]. The design is unusually demanding because it tracks the same tumor through space and time, which is precisely where classification instability tends to hide.

Alongside that primary data, a January 2025 review synthesized the broader literature and reached a candid conclusion: AI tools are "stumped" by the HER2-low boundary in much the way pathologists are [2]. And a forward-looking 2026 state-of-the-field review, published in mid-2025, situates all of this within the wider trajectory of digital and computational pathology, adoption standards, clinical integration, and the specific gap between AI tools that carry regulatory clearance and their validated performance at the HER2-low/ultralow threshold [3]. Taken together, these aren't announcements of a solved problem. They're a shared acknowledgment that the hardest part of HER2 scoring may be irreducible with current approaches.

The diagnostic problem

To understand why this matters, you have to appreciate how much diagnostic weight now rests on a distinction the original HER2 assay was never designed to make.

HER2 (ERBB2) immunohistochemistry in breast cancer was built to sort tumors into a clinically actionable dichotomy: positive (3+, or 2+ with amplification confirmed by in situ hybridization) versus negative. Everything below the amplification line, IHC 0, 1+, and non-amplified 2+, was lumped together as "HER2-negative" because none of it predicted benefit from the classic HER2-targeted agents. The scoring anchors, the training, the proficiency testing all evolved around that binary.

The arrival of antibody, drug conjugates changed the stakes. Suddenly the tumors formerly dismissed as negative were resolved into meaningful strata, HER2-low (IHC 1+, or 2+ without amplification) [6] and, more recently, HER2-ultralow (IHC 0 with faint, incomplete membrane staining in ≤10% of cells, above a true, staining-free 0) [5][7]. This matters for eligibility considerations around an entire drug class rather than any single, patient-directed regimen. But the assay is being asked to draw a line it wasn't calibrated to draw: the boundary between 0 and 1+, and between 0 and ultralow, hinges on faint, incomplete, sometimes barely perceptible membrane staining.

That's an unforgiving place to demand reproducibility. Inter-observer agreement, solid at the 3+ end, degrades sharply at the low end. Two competent pathologists can look at the same faint staining and land on different sides of a boundary that now carries therapeutic weight. This is the reproducibility problem computational analysis was supposed to help solve, and it's the substrate on which all three developments sit.

The evidence

Here's where the evidence gets interesting, and sobering.

The 47-case study demonstrates two things at once. First, HER2-low and HER2-ultralow status is unstable, it converts across time points and tissue types, so a core biopsy, its matched resection, and a distant metastasis from the same patient can disagree with one another. This instability is well documented at the low end: roughly 40% of cases switch between IHC 0 and HER2-low on reassessment [5]. A June 2025 matched-specimen study tracking the same tumors across core, resection, and metastasis reports the same direction of travel [1]. That's a biological signal, not a technical artifact, and AI-assisted digital pathology captured these changes alongside conventional microscopy. Second, and this is the key caveat, the study found that AI does not resolve the classification instability inherent to the scoring boundary itself [1]. It can measure the moving target faithfully. It cannot make the target stop moving.

A January 2025 review reaches the same conclusion from the technical side, and the vault's own primary evidence bears it out. Computational image analysis improves quantitative membrane scoring, it can count stained cells and grade completeness more consistently than an eye estimating percentages under time pressure, the same analytical-validation gain seen in automated scoring of other breast IHC markers, where an algorithm reproduces a reference read with an intraclass correlation of 0.938 [8]. But that quantitative gain doesn't translate into eliminating the core reproducibility problem at the IHC 0-versus-1+ threshold, where AI assistance lifts agreement only to 87.4% [4][2]. The evidence suggests the bottleneck isn't the algorithm's measurement precision; it's the absence of standardized training datasets and consensus scoring anchors to teach the model where the line falls. If the ground truth is contested, a model trained on contested labels inherits the contest.

The 2026 state-of-the-field review adds the systems-level view: AI-driven HER2 scoring has advanced, and clinical integration milestones, including HER2-low use cases, are real, but a gap persists between tools that carry regulatory marking and their demonstrated performance at exactly the clinically critical boundary [3].

The honest status

So what's the honest status?

Every development discussed here is investigational in the sense that matters most, none has been shown to resolve the ultralow reproducibility problem, and the reviews are explicit that a gap remains between CE-marked or FDA-cleared AI HER2 tools and validated performance at the low/ultralow threshold [3]. Clearance for a HER2 scoring workflow is not the same as validated reproducibility at the 0-versus-ultralow call. Readers should hold those two facts apart.

The open questions are concrete. Can consensus scoring anchors be established for ultralow, and can training datasets be standardized enough that models trained in one lab generalize to another [2]? How should the field handle the biological reality that HER2-low status changes across a patient's disease course, so that no single measurement, human or machine, is the whole story [1]? And what validation bar should regulators and laboratories demand specifically at the boundary that carries therapeutic consequence, rather than at aggregate accuracy that's dominated by easy 0s and easy 3+s?

The measured read is this: computational HER2 analysis is a real advance in consistency of measurement, and a real disappointment as a solution to the boundary problem, because the boundary problem was never only about measurement. It was about where we've agreed to draw the line, and we haven't fully agreed yet. The algorithm, it turns out, is waiting on us.

---

References

1. Changes in HER2-low and HER2-ultralow status in 47 advanced breast carcinoma core biopsies, matching surgical specimens, and their distant metastases assessed by conventional light microscopy, digital pathology, and artificial intelligence. Breast Cancer Res Treat. 2025. PMCID: PMC12396990. [full author list/volume to be completed from the record before publication].

2. HER2-Low Breast Cancer at the Interface of Pathology and Technology: Toward Precision Management. Biomedicines (MDPI) 2025;14(1):49. DOI: 10.3390/biomedicines14010049. [full author list to be completed from the MDPI record].

3. What's new in digital and computational pathology 2026: advances in adoption, standards, AI technologies, and clinical integration. 2025. PMCID: PMC13183467. [full author/journal detail to be completed from the record].

4. Krishnamurthy S, et al. Fully automated artificial intelligence, based HER2 immunohistochemistry scoring in breast cancer. JCO Precis Oncol. 2024. PMID: 39393036. doi:10.1200/PO.24.00353.

5. Wolff AC, Hammond MEH, Allison KH, et al. Human epidermal growth factor receptor 2 testing in breast cancer: ASCO, College of American Pathologists guideline update. Arch Pathol Lab Med / J Clin Oncol. 2023. PMID: 37303228. doi:10.1200/JCO.22.02864.

6. Modi S, Jacot W, Yamashita T, et al. Trastuzumab deruxtecan in previously treated HER2-low advanced breast cancer (DESTINY-Breast04). N Engl J Med. 2022. PMID: 35665782. doi:10.1056/NEJMoa2202914.

7. Bardia A, Hu X, Dent R, et al. Trastuzumab deruxtecan after endocrine therapy in metastatic HER2-low/HER2-ultralow breast cancer (DESTINY-Breast06). N Engl J Med. 2024. PMID: 39282896. doi:10.1056/NEJMoa2407086.

8. Boyaci C, et al. Automated Ki-67 scoring in ER-positive breast cancer (International Ki67 in Breast Cancer Working Group). Biomolecules. 2021. PMID: 34827609. doi:10.3390/biom11111612.