固体瘤 HER2 检测一本通：CAP 指南册、器官系统评分与低/超低表达One for all: HER2 guidebook for solid tumors

Karen Titus

Why a HER2 guidebook is needed

Though it may test the memory of many, there once was a time when HER2 immunohistochemistry was a relatively uncomplicated matter. Back in September 1998, the FDA approved both trastuzumab (Herceptin) and the first companion diagnostic assay, HercepTest, which was used to determine whether patients with metastatic breast cancer could receive the new drug.

A strong positive (3+) result was a green light for treating patients; an equivocal (2+) result would mean reflexing to in situ hybridization to look for gene amplification.

Groundbreaking? Yes. Heady and exciting? Yes.

Naturally, there were many questions as pathologists sorted through the intricacies of testing over the next several years. Nevertheless, it was a relatively easy hike from test to treatment.

And the HER2 landscape these days?

“Now life is not as simple anymore,” says Gulisa Turashvili, MD, PhD, who oversaw the creation of a new guidebook to HER2 testing in solid tumors (Turashvili G, et al. Arch Pathol Lab Med. Published online June 19, 2026. doi:10.5858/arpa.2025-0593-CP). The advent of antibody-drug conjugates in particular, including T-DXd, now has pathologists and their clinical colleagues traversing a much more demanding trail.

Hence the new guidebook, says Dr. Turashvili, subspecialty director of gynecologic pathology, Mass General Brigham, and associate professor of pathology, Department of Pathology, Massachusetts General Hospital and Harvard Medical School. It’s the work—and first publication—of the Biomarker Reporting Project Team, composed of members of a number of CAP committees (Cancer, Pathology Electronic Reporting, Immunohistochemistry, Center Guideline, Molecular Oncology). Dr. Turashvili is co-vice chair of the Cancer Committee.

The team evaluated current evidence for HER2 testing, interpretation, and reporting in multiple organ systems. Of 11 currently active CAP biomarker reporting templates, seven (breast, colon/rectum, gastric, gastrointestinal stromal tumor, gynecologic, head/neck, and non-small cell lung cancer) contain elements for capturing HER2 status. As the authors note, these wide-ranging elements make it difficult for pathologists to accommodate different scoring systems, direct appropriate therapy, and provide clinical colleagues with clear and concise HER2 reporting.

Or, to put it more plainly, “It’s all very challenging,” says Dr. Turashvili. “For all of us.”

Given the complexities of HER2 testing, the number of organ systems that are implicated, and the fast pace at which knowledge has evolved (case in point: the groundswell of interest in identifying HER2 low and ultralow status by IHC), Dr. Turashvili suspects that the new guidebook will be of use to nearly everyone who is involved in biomarker testing or solid tumors, especially practicing anatomic and surgical pathologists, whether in an academic center or in community practice. “Most of us do get requests for biomarker testing, including HER2.”

Clinicians could also find the guidebook useful.

“Quite often I come across questions, or I become part of conversations, where it’s obvious to me that there’s a misunderstanding or lack of understanding of how HER2 assessment is actually done on solid tumors,” she says. Clinical colleagues might not comprehend the various criteria pathologists apply to different tumor sections or tumor types to determine whether the tumor is HER2 positive or negative or if it has low HER2 expression.

“You have to be careful about scoring criteria or scoring guidelines that you apply to each tumor type before you come up with a final result,” she says, adding that this can be somewhat confusing for pathologists as well.

As she oversaw the development of the guidebook, she says even she encountered a few surprises. “I knew this was quite complicated, but I may have slightly underestimated how challenging it would be to put all of this together,” she says with a laugh. Sometimes there was similar information, but presented differently. In other cases, the information was subtly different. And she also gained insight into HER2 testing outside her specialty. “Because I only practice breast and gynecologic pathology.”

Nor does she want to overlook the value of the guidebook for trainees and fellows in various subspecialties. “It’s hard to think of anyone who has anything to do with biomarker assessment who wouldn’t benefit from this guidebook,” she says.

The complexity of organ-system scoring

Assessing HER2 for the various organ systems is more like a passeggiata, leaving physicians to interact with an overlapping mix of sample requirements and testing methodologies, recommended scoring criteria for IHC and ISH, and various definitions of positivity, in a variety of clinical settings and with ever-evolving targeted therapies.

Little wonder physicians grapple with questions and misunderstandings.

Determining the appropriate targeted therapy for each patient is the main question, Dr. Turashvili says, adding that the advent of ADCs means pathologists will need to work up cases using different criteria depending on what treatment the clinician is considering.

“The devil really is in the details,” she says. The guidebook doesn’t shy away from these finer points. For the organ systems covered in the publication, the authors include sections on the historical overview and current indications for HER2 testing, samples to test, scoring algorithms for IHC and ISH, disease-specific considerations, and future directions. (The guidebook also discusses HER2 testing in cytology specimens, as well as emerging opportunities and challenges, including digital pathology in HER2 ISH, and DNA and RNA sequencing and polymerase chain reaction-based detection of HER2 gene amplification.)

Dr. Turashvili delves into the details using the example of gynecologic tumors. A clinical trial that opened in 2011 included a specific group of patients with endometrial serous carcinoma; patients were treated with a combination of trastuzumab as anti-HER2 targeted therapy and chemotherapy, “with remarkable improvements in progression-free and overall survival,” she says.

The trial used modified breast scoring criteria to enroll the patients. “So now we have this so-called gynecologic cancer-specific criteria, or endometrial-specific criteria, that we apply to endometrial cancer patients to determine eligibility for trastuzumab,” Dr. Turashvili says.

However, “These criteria are not the same as the current breast criteria.” Without knowing the appropriate specific cutoffs, pathologists can’t determine HER2 positivity by IHC. “But it’s probably sufficient to say that the cutoff to determine HER2 positivity by immunohistochemistry is higher than what we currently use for breast cancer—basically, it’s 30 percent versus 10 percent for HER2 protein overexpression.”

More recently, the DESTINY-PanTumor02 trial essentially turned up the house lights, showing that T-DXd had treatment value for a variety of solid tumor types, including ovarian, endometrial, and cervical cancers. As use of T-DXd has expanded, so has the need to fine-tune scoring criteria, with gastric/gastroesophageal criteria applied in a variety of settings.

For endometrial cancers, “you have two sets of gastric criteria,” Dr. Turashvili says. “Different criteria for surgical specimens versus biopsies. It’s still gastric criteria, but depending on what type of specimen you’re looking at, you’re going to modify the criteria slightly. And these criteria are different from the [aforementioned] breast criteria or from the endometrial-specific criteria.”

Another frequent misunderstanding can be traced back to the earliest days of HER2 testing in breast cancer, when IHC 2+ tumors would be reflexed to ISH.

As Dr. Turashvili explains, “The huge difference with gastric criteria—and this is where people get confused—is that in the DESTINY-PanTumor trial, the gastric criteria did not require ISH testing. What that means for us as pathologists is that if one applies the gastric criteria for the solid tumors—including endometrial, ovarian, cervical—you do not need to do in situ hybridization for 2+ tumors, because you’re going to follow the trial criteria.”

The shift “is really hard to wrap your head around,” she says, given that it’s second nature for breast pathologists to follow a 2+ IHC with ISH to determine if a patient qualifies for treatment. “T-DXd, and ADCs, and the DESTINY-PanTumor trial, changed everything.”

Such upheaval drives home another important point, one she and her coauthors make clear in the guidebook. Says Dr. Turashvili: “When one reports HER2 results, for whatever tumor type, it is essential to specify what criteria they applied to come up with a score.” Simply reporting a result as positive or negative—or even specifying a score—“is no longer enough because everything is now much more complicated.”

Despite the many differences outlined in the guidebook, HER2 testing also offers some common ground, Dr. Turashvili says.

Referring to the DESTINY-PanTumor trial and T-DXd therapy, she says, “All these tumors can be scored by gastric criteria.” Within the gynecologic tract, for example, “If I look at ovarian cancer versus cervical cancer, if I know that the clinician is considering T-DXd, I don’t have to worry about applying different scoring systems,” she says. Likewise, if a pathologist is looking at a urothelial cancer and knows the patient might be treated with T-DXd, “they would apply the exact same gastric criteria that would be applied to ovarian cancer.”

In short, the crux of the matter is the intended therapy. If a case involves a solid tumor and T-DXd, she says, “Then I use gastric criteria and just look at IHC. But again, keep in mind that there are slightly different criteria for the surgical specimens versus biopsies.”

HER2 low, HER2 ultralow, and DESTINY-PanTumor

The DESTINY-PanTumor trial and growing use of T-DXd have opened up another avenue of discussion: What are the implications of HER2 low and HER2 ultralow? These days, it matters.

Prior to this, when the focus was primarily on breast cancer and trastuzumab, tumors that were scored as either 1+ by IHC or 2+ by IHC and negative for gene amplification by ISH were deemed HER2 negative (now considered HER2 low); those patients received a different treatment (i.e. not trastuzumab).

And scores of 0 by IHC, which traditionally were seen as a negative result, meant patients weren’t considered eligible for any type of HER2-targeted therapies. But with the advent of ADCs, it has become important to tease out the specifics of the lower scores. The CAP Cancer Committee modified the breast biomarker reporting template to account for those low levels of HER2 expression. Previously, says Dr. Turashvili, a score of 0 meant there was less than 10 percent of tumor cells with barely perceptible incomplete staining or no staining at all. In 2025, they split the score of 0 into two categories: score 0 defined as negative with no membrane staining, and score 0+ denoting negative with membranous staining.

“Five years ago, if someone told you this is what we would be doing now, you’d say, That doesn’t make any sense. Why would you need to know that? It’s score 0,” she says. But current treatment options make the distinction important.

It’s important to realize, Dr. Turashvili continues, that low levels of expression can be subject to interobserver variability. “And more importantly, you may get different results with different immunohistochemical protocols,” given the various HER2 antibodies, staining platforms, and detection systems. And, she notes, “just different labs. Even if you use the same antibody, the same detection system, and the same platform, and you run HER2 IHC on the same tissue block, you may still get different results when it comes to very low levels of HER2 protein in the tissue.” Quite simply, HER2 IHC is not the ideal test for identifying extremely low levels of HER2 in tissue. “But it’s the only clinically available test we have right now. It’s a huge limitation.”

Dr. Turashvili refers to this issue as one of two elephants in the room.

HER2 IHC was optimized and validated for identifying tumors with IHC 3+ scores, but its use, obviously, has expanded to cases with low levels of expression. “We don’t talk about it very much,” she says—her own actions notwithstanding. “Oh, I talk about it every chance I get,” she says. “People need to understand this is not a perfect test, and it may not even be an appropriate test when it comes to low levels of HER2 expression.”

Other tumor types and specimen-handling challenges

Breast tumors are obviously well studied and well known—given their historical importance, they might be considered the George Washingtons of HER2-related tumors. The various gynecologic tumors and GI carcinomas are becoming better understood as well. But the other solid tumors with HER2 implications, while covered in the guidebook, are less appreciated, in no small part because data on these are limited.

Nevertheless, Dr. Turashvili hopes the guidebook will help expand physicians’ perceptions.

“If patients don’t have any other treatment options,” she suggests, “one could look at the guidebook and learn that urothelial cancers can be treated with HER2-targeted therapy.” That’s also true for testicular cancer, prostate cancer, renal cancer, head and neck cancers, and lung cancer. The information is summarized in Table 1 of the guidebook.

“What some people don’t realize,” she says, “especially if they don’t practice all these different subspecialty areas, is that even with rare tumor types,” HER2-targeted therapies might be options for patients beyond first-line therapies.

As the guidebook makes clear, each specimen has its own challenges, she says. Which leads her to the second elephant in the pathology laboratory.

This particular pachyderm is tied to fixation time and cold ischemic time, preanalytical variables that have long been tracked in breast cancer specimens. Documenting those variables—ischemic time of less than 60 minutes and optimal fixation time of six to 72 hours—is critical (as well as a requirement for CAP-accredited laboratories), with the information included in pathology reports, she notes.

“However, we don’t track fixation time or cold ischemic time for gynecologic specimens, or any other types of specimens besides breast,” she says. “It’s not something we talk about too much, but it’s logical to think it would be important for other specimens. But I don’t have any data to support that logical assumption.”

While underfixation is relatively easy to appreciate under the microscope, overfixation is not. That could cause problems, she says. “If you’re doing stains and everything comes back negative, and it doesn’t make any sense, one might wonder if it’s overfixation.” But when trying to discern between 1+, 2+, and 3+ results, the impact of overfixation may not be obvious.

Challenges of unified biomarker reporting

Like the proverbial tree in the forest, a test result requires a clear report to achieve meaning. The guidebook also explores the challenges of designing a unified biomarker reporting format to cover alternative systems.

Dr. Turashvili leads the CAP Cancer Biomarker Reporting Project Team, which is tasked with optimizing biomarker reporting for the cancer protocols.

“One of the issues we’ve been discussing lately is harmonization and standardization of biomarker reporting, including HER2,” she says. Indeed, HER2 is a primary example, “because it’s probably one of the most tested single biomarkers for different tumor types.”

“In theory, or ideally, one would create a single HER2 reporting template that can be used by different subspecialty pathologists for different tumor types.”

But theory and reality can be awkward dance partners. “We don’t always use the same scoring system for all these different tumor types,” she says. Likewise, there is often a set or panel of biomarkers that one would do for a specific tumor type, including but not restricted to HER2. An endometrial cancer case, for example, would include p53, mismatch repair proteins, ER, and so on. “It’s a specific panel of stains that doesn’t necessarily work for breast cancer, because we don’t usually do MMR or p53 for breast cancer.”

And herein lies the difficulty: How can pathologists use a single HER2 reporting template when they’re also adding other markers for a specific tumor type?

Even if there were a single synoptic report, Dr. Turashvili adds, other problems lie in wait. It’s not unusual to use numerous criteria—a colorectal cancer specimen might involve use of both gastric and HERACLES criteria, for example—depending on what therapy the clinician is considering. What would be the best way to incorporate different criteria?

These are open questions, she says. “We are working on it. I don’t know what the answer is right now, or what we’re going to come up with.” One possibility could be using an internal template, with options for using different scoring systems. “Then, in theory, we could build tumor-specific synoptic reports where you take the section that is applicable to your tumor type. But even that becomes challenging,” she acknowledges.

Is it a realistic goal to develop a single synoptic template that works for everyone all the time?

“There probably is a way,” she muses. “And there is a need for it. It’s good to have a harmonized, standardized system, but it should not be at the expense of losing all the subspecialty-specific, subtle details that need to be incorporated into the report. I’m optimistic we will come up with something that works.”

A first step and future updates

The guidebook is, like the first moon landing, a giant leap. It’s the first product developed by the Biomarker Reporting Project Team; Dr. Turashvili suspects this may be the first publication that summarizes HER2 reporting for different solid tumor types.

Future iterations will be subject to the ebb and flow of research advancements, which are unpredictable.

“Usually it takes time between one big trial and the next,” she says. “But you never know. Things have been moving really fast lately.” It’s possible, if there is another groundbreaking trial similar to DESTINY-PanTumor, “we would absolutely need to update this.”

Apart from that, “There is no plan to update it regularly,” she says, but rather, “as soon as there are important changes that make what we say in this guidebook incorrect.”

Karen Titus is CAP TODAY contributing editor and co-managing editor.

Amy Carpenter is a writer for CAP TODAY.