The landscape of dynamic tumor immunophenotyping on neoadjuvant chemotherapy combined with trastuzumab for the treatment of HER2-positive breast cancer.

BACKGROUND: For HER2-positive breast cancer (BC), patients who achieve pathologic complete response (pCR) are predicted to have better clinical outcomes. With the advent of the era of neoadjuvant chemotherapy (NAC) combined with targeted therapy, the pCR rate of HER2-positive BC has increased significantly, but about 50% of these patients still do not respond to targeted therapy. The dynamic changes in the tumor microenvironment (TME) may crucially influence pCR outcomes in HER2-positive BC. However, there is no specific immunophenotyping study that correlates with the efficacy during treatment. METHODS: We conducted a comprehensive spatial expression profiling analysis on 94 ROIs procured from 28 HER2-positive BC patients, encompassing both their initial diagnostic stage and postoperative status. Leveraging DSP techniques, we precisely mapped and quantified the TME at these critical time points for patients. RESULTS: Notably, we classified specimens into four groups: IM1, IM2, TM1, and TM2. IM1 and TM1, characterized by CD4⁺ naïve T cells and naïve B cells, showed higher pCR rates, whereas IM2 and TM2 exhibited lower rates. In addition, baseline IM1 patients who were classified as TM1 postoperatively achieved a 100% pCR rate, while baseline IM2 patients who were classified as TM2 postoperatively had a poor pCR rate of only 10%. The immunophenotypes defined by spatial transcriptomics were reproducible in both TCGA and single-cell datasets, and spatial analysis further revealed structured cellular transitions and functional heterogeneity at the tumor-immune interface. CONCLUSION: Our dynamic immunophenotyping can provide prediction of the efficacy for NAC with trastuzumab in HER2-positive BC patients, and provide a basis for the selection of subsequent treatment regimens.