Discrepancies in Hormone Receptor and HER2 Expression between Malignant Serous Effusions and Paired Tissues from Primary or Recurrent Breast Cancers

Abstract

Introduction: Immunohistochemistry (IHC) for the estrogen receptor (ER), progesterone receptor (PR), and human epidermal growth factor receptor 2 (HER2) biomarkers has prognostic and therapeutic value in breast cancer. This study aimed to compare the expression of ER, PR, and HER2 between paired malignant effusions and tissue samples of breast cancer. Methods: Our electronic archive was searched for all effusions diagnosed as breast carcinomas within a pre-defined period (January 2018–October 2021). Next, their cell blocks (CBs) were subjected to ER, PR, HER2 IHC, or in situ hybridization, in addition to EGFR IHC. The expression of hormone receptors (HRs) and HER2 was subsequently compared between tissue and effusion cytology samples derived from the same patients. Results: Only 2/76 (2.6%) of the breast cancer patients analyzed showed a malignant effusion at their initial presentation. ER, PR, and HER2 discordance rates between paired malignant effusions and tissue samples obtained at initial diagnosis were 24.3% (17/73), 40.8% (29/71), and 9.1% (6/66), respectively. The HR–/HER2– status was found more often at effusions compared to paired tissue biopsies obtained at initial diagnosis (30/70 vs. 17/70; p < 0.001). In addition, the HR–/HER2– status was significantly associated with an earlier development of a malignant effusion, when found at initial diagnosis (p < 0.001; log-rank test), first recurrence/metastasis (either solid or effusion) (p = 0.012; log-rank test), effusion samples (p = 0.007; log-rank test), and any tumor sample obtained (p = 0.009; log-rank test). Lastly, EGFR overexpression in the HR–/HER2– effusion samples was significantly associated with a shorter post-effusion survival (p = 0.019; log-rank test). Conclusion: Serous effusion cytology provides high-quality material for ancillary techniques, especially when CBs are prepared, reflecting cancer heterogeneity.