Pathologic landscape of HER2-low in gastroesophageal adenocarcinoma: real-world data

Background. Human epidermal growth factor receptor 2 (HER2) is the first molecular biomarker which has been exploited for advanced gastroesophageal cancers’ targeted therapy. On the basis of the results from the ToGA trial, the anti-HER2 monoclonal antibody trastuzumab combined with chemotherapy is routinely used as the first-line therapy for HER2-positive advanced gastroesophageal cancers. The DESTINY-Gastric01 trial showed that a novel HER2-targeted antibody-drug conjugate, Trastuzumab-deruxtecan, in addition of being valuable in overcoming the problem of HER2 intra-tumour heterogeneity, has proved to be effective also in those cancers which have a low HER2-expression rate (HER2-low disease, defined as HER2 immunohistochemistry 1+ or 2+ without gene amplification confirmed by in situ hybridization), thus paving the way for novel therapeutic scenarios. Aim of the study. The purpose of this study was to evaluate the prevalence of HER2-low expression in a large real-world and multi-institutional series of cases of gastroesophageal adenocarcinomas. In addition to the prevalence analysis, the study also aimed to evaluate the correlation between this low expression rate with a series of clinical and histopathological features of gastroesophageal cancers. Materials and methods. We retrospectively evaluated a total of 1.210 formalin-fixed paraffin-embedded samples of gastroesophageal adenocarcinomas which were analyzed by immunohistochemistry for HER2 protein expression in eight Italian surgical pathology units in the period between January 2018 and June 2022. We assessed the prevalence of HER2-low (i.e., HER2 1+ and HER2 2+ without amplification) in the cohort of available samples. Each specimen was also evaluated and categorized considering different features, such as the type of specimen (surgical or biopsy), number of biopsy fragments, tumor localization, histotype according to the WHO 2019 criteria, grading, histopathological characteristics according to Lauren and Ming classification systems. It was considered also the year and the center where the specimen was collected. Information regarding staging, neoadjuvant therapy and other biomarkers’ status (PD-L1, dMMR/MSI status, EBER) was also collected from the pathology reports. Results. HER2 status could be assessed in 1.189/1.210 cases. Among the 1.189 assessable cases, 710 (59,7%) were HER2 0, 217 (18,3%) were HER2 1+, 120 (10,1%) were not amplified HER2 2+, 41 (3,4%) were amplified HER2 2+, and 101 (8,5%) were HER2 3+. The prevalence of HER2-low was 28,3% (95% CI 25,8 to 31,0%) overall, and was higher in biopsy specimens (34,9%, 95% CI 31,2 to 38,8%) rather than in surgical resection specimens (21,0%, 95% CI 17,7 to 24,6%) (p<0,0001). Moreover, HER2-low prevalence ranged from 19,1 to 40,6% among centers (p=0,0005), and from 26,6 to 40,6% according to the clone of the antibody used for the immunohistochemical staining (p=0,01). It was pointed out also that HER2-low prevalence was lower in pure signet-ring carcinomas (p=0,07). Conclusion. In the light of the promising activity of trastuzumab-deruxtecan in advanced HER2-low expressing gastroesophageal cancers and in the wake of the knowledge available on HER2-low breast cancers, the new “HER2-low” category in gastroesophageal cancers may be the starting point toward a reconsideration of the world of HER2 expressing cancers. However, this work shows how the expansion of the HER2 spectrum might raise problems in reproducibility, especially in biopsy specimens, decreasing inter-laboratory and inter-observer concordance. This may make it necessary to better define the characterization of HER2-low category.