The newly identified category of “HER2-low” breast cancer has raised many new issues in this malignancy now that the DESTINY Breast-041 trial determined that fam-trastuzumab deruxtecan-nxki (T-DXd) may effectively treat this tumor. Among the issues are challenges in accurately identifying just which tumors have low HER2 expression, according to David Rimm, MD, PhD, the Anthony N. Brady Professor of Pathology at Yale University, New Haven, Connecticut.
David Rimm, MD, PhD
Limitations of Immunohistochemistry
At the 2022 San Antonio Breast Cancer Symposium, Dr. Rimm offered what he called a “progressive pathologist’s” perspective on issues surrounding HER2.2 “Instead of having a cut point between HER2-positive and HER2-negative, we now have a HER2-low category that we have to figure out how to work with,” he said.
So-called HER2-low tumors have HER2 proteins on their cell surface but in quantities insufficient to classify them as HER2-positive. According to a growing body of data, the determination of HER2-low status by immunohistochemistry (IHC), the standard assay approach, is imperfect. “The current IHC assay for HER2 is not reproducible between pathologists, since it is not designed for the low-HER2 dynamic range. This fact makes it difficult to conclusively determine whether HER2-low is a separate entity if IHC is used to stratify study groups,” he said.
Lack of reproducibility was, indeed, shown in a recent study by Fernandez et al.3 Dr. Rimm and his team examined the suitability of the current IHC assay to select patients with low HER2 positivity for treatment with T-DXd. Findings from approximately 1,400 laboratories suggested that scoring accuracy was poor.
“We found there was a discordance of 41% in distinguishing HER2 0 from HER2 1 or higher, but a discordance of 11% between HER2 3 vs not 3. We only get it right about 60% of
the time,” Dr. Rimm shared. The data also showed, he added, that the “most liberal” of the 18 pathologists called 40% of samples HER2 0, and the “most conservative pathologists” called 20% HER2 0.
New Statistical Approach: Comparing Study Findings
To further evaluate interobserver agreement on identifying HER2-low tumors, Dr. Rimm applied a new statistical approach called an ONEST (observers needed to evaluate a subjective test) plot, which maps the overall percent agreement vs the number of pathologists in a group of readers. “The analysis showed we are pretty darn good at determining estrogen receptor–positive from estrogen receptor–negative status [for comparison], and for HER2 3+ vs not 3+ we do really well, but for HER2 0 vs not 0, you see a very wide range, where two pathologists agree only 50% of the time,” he explained.
He noted that the 60% concordance found in the Fernandez study was approximately 20% less than the figures reported in several posters at an SABCS session on HER2-low disease. Thus, these findings suggest that pathologists may often achieve an acceptable level of accuracy:
From DESTINY-Breast04: There was 77% agreement between historical and central HER2-low status using the Ventana PATHWAY 4B5 IHC assay and Ventana INFORM HER2 dual ISH assay.4
From Viale et al: In 781 samples, the overall percentage agreement between rescored and historical HER2 scores was 81%; agreement was numerically greater for HER2-low than HER2 IHC 0.5
From Rüschoff et al: Among 80 pathologists, the overall score concordance with a new category of HER2-low was above the 80% overall rater agreement benchmark for both the Ventana PATHWAY 4B5 and Dako HercepTest HER2 assays and was higher than previous reported by Fernandez et al in Dr. Rimm’s group.3,6
According to Dr. Rimm, those findings were overly optimistic and potentially flawed, and as such they could ultimately affect patient care. “A key consideration is that all data published to date depend on a flawed premise. Whether you think HER2-low is a separate entity or not, remember that anyone who uses IHC 0 vs IHC 1+ to dichotomize their data is subject to error,” he maintained.
Next Steps for HER2-Low Identification
Dr. Rimm argued that with current assays it is not possible to use IHC assays to accurately adjudicate HER2 0. “I would argue that most oncologists would not accept an ordinal value for sodium, magnesium, or glucose, and neither should they accept an ordinal value for HER2,” he said. “I propose that oncologists should request us pathologists to be more accurate…. We need to give you detailed information or actual numbers…. I think it’s time to go digital … to tell which patients truly have HER2 present and which patients truly do not.
“A number of labs are working on this,” Dr. Rimm said. Dr. Rimm has personally been involved in researching the Automated Quantitative Analysis, or AQUA™ (HistoRX), method of quantitative immunofluorescence for testing a range of antibody concentrations; this may help to maximize sensitivity within the lower range of HER2 expression. AQUA uses a cell line microarray with HER2 protein measured by mass spectrometry to determine the amount of HER2 protein in units of attomoles (10–18 moles/mm2). The assay has the correct dynamic range, he added, “so we are not weighing mice on a scale built for elephants.”
“This information or other data from a new assay with the correct dynamic range may lead to a better understanding of HER2-low cancer,” Dr. Rimm said. “I encourage you all to measure, not read, HER2.”
DISCLOSURE: Dr. Rimm has served as a consultant or advisor to AstraZeneca, Amgen, Cell Signaling Technology, Cepheid, Danaher, Immunogen, Janssen, Merck, NextCure, PAIGE.AI, Regeneron, Sanofi, and Ventana/Roche. Research in Dr. Rimm’s lab is currently funded or was funded in the past by Amgen, Cepheid, NavigateBP, NextCure, and Konica/Minolta.
REFERENCES
1. Modi S, Jacot W, Yamashita T, et al: Trastuzumab deruxtecan in previously treated HER2-low advanced breast cancer. N Engl J Med 387:9-20, 2022.
2. Rimm D: HER2 low: A pathologist’s perspective. 2022 San Antonio Breast Cancer Symposium. Special Session. Presented December 7, 2022.
3. Fernandez AI, Liu M, Bellizzi A, et al: Examination of low ERBB2 protein expression in breast cancer tissue. JAMA Oncol 8:1-4, 2022.
4. Prat A, Modi S, Tsurutani J, et al: Determination of HER2-low status in tumors of patients with unresectable and/or metastatic breast cancer in DESTINY-Breast 04. 2022 San Antonio Breast Cancer Symposium. Abstract HER2-18. Presented December 7, 2022.
5. Viale G, Basik M, Nikikura N, et al: Retrospective study to estimate the prevalence and describe the clinicopathological characteristics, treatment patterns, and outcomes of HER2-low breast cancer. 2022 San Antonio Breast Cancer Symposium. Abstract HER2-15. Presented December 7, 2022.
6. Rüschoff J, Penner A, Ellis IO, et al: Proficiency assessment of HER2-low breast cancer scoring with the Ventana PATHWAY 4B5 and Dako HercepTest HER2 assays and the impact of pathologist training. 2022 San Antonio Breast Cancer Symposium. Abstract HER2-13. Presented December 7, 2022.
