Purpose: Somatic alteration of phosphatidylinositol-4,5-bisphosphate 3-kinase catalytic subunit alpha (PIK3CA) is a crucial therapeutic target in breast cancer (BC) and PI3Kα-specific inhibitor Alpelisib has been used in clinics. This study investigates the PIK3CA alterations in Chinese and Caucasians BC patients for the purpose of selecting anti-PI3K therapy. Methods: The molecular profile of the PIK3CA gene was analyzed in 412 Chinese patients with untreated invasive BC using a 540 gene next-generation sequencing panel. The results were compared with data of the Caucasian BC patients in The Cancer Genome Atlas (TCGA-white). Results: PIK3CA alterations were frequently found in BC of estrogen receptor (ER) positive (49.3%, p = 0.024), low ki67 proliferation index (58.3%, p = 0.007) and low pathological grade (grade I/II/III 80%, 53.4%, 35.9%, p < 0.001). Compared to TCGA-white, Chinese BC patients had a higher alteration frequency (45.6% vs. 34.7%, p < 0.001) with larger proportion of p.H1047R mutation among three common mutation sites (p.E545K, p.E542K and p.H1047R) (66.1% vs. 43.7%, p = 0.01). Across four molecular subtypes, ER + /human epidermal growth factor receptor 2 positive (HER2 +) tumors harbored the most PIK3CA alterations (51.6%), while ER-/HER2- harbored the least alteration (30.0%) but the most copy number amplification (19.05%). Conclusion: PIK3CA alterations prevail in Chinese BC patients and have different molecular features compared to that of Caucasians. The results provide precise annotations of PIK3CA genomic alterations of Chinese in the context of application of PIK3CA inhibitor.

Checkpoint inhibitor therapy has become increasingly important and has been endorsed as a treatment regimen in breast cancer. But benefits were limited to a small proportion of patients. We aimed to develop an improved signature on the basis of immune genes for detection of potential benefit from immunotherapy. Gene expression data of patients with breast cancer initially extracted from The Cancer Genome Atlas were analyzed. Ten genes were selected from the interaction of differentially expressed genes as well as immune-related genes to develop a survival signature. We compared the high-risk and low-risk groups by gene set enrichment analysis, immune infiltration, checkpoint molecule expression and immunophenoscore. Ten genes were extracted from interactions of differentially expressed and immune-related genes. The immune risk score was determined on the basis of the Cox regression coefficient of hub genes and validated with the GSE96058 dataset. Immune cell infiltrates, including CD8 + T cells, plasma cells, follicular helper T cells, CD4 + memory T cells, M1 macrophages, regulatory T cells and resting NK cells, were more highly infiltrated in the high-risk group as compared to the low-risk group. Checkpoint molecules, including CTLA-4, PD-L1, TIM-3, VISTA, ICOS, PD-1, and PD-L2, were expressed at markedly lower levels in the high-risk group as compared to the low-risk group. Immunophenoscores, as a surrogate of response to immune checkpoint therapy, was observed significant lower in the high-risk group. The 10-gene prognostic signature could identify patients’ survival and was correlated with the biomarkers of immune checkpoint inhibitor therapy, which may guide precise therapeutic decisions in clinical practice.

This pilot study evaluated the feasibility of a novel technique combining ultrasound-guided and indocyanine green (ICG) intra-tumoral injection with robot-assisted surgery to enable accurate local excision (lumpectomy) for breast cancer. Robot-assisted breast surgery has been used for nipple-sparing mastectomy for breast cancer; however, there are few published data on robot-assisted lumpectomy. A novel surgical technique combined intraoperative ultrasound and indocyanine green (ICG) injections was developed to delineate narrow surgical excisions in 63 breast cancer patients by a single surgeon with precisely timed robotic-assisted excision. The surgical margins were assessed pathological with intraoperative frozen section examination. ICG-marked tumor boundaries were identified with clips and later evaluated using permanent sections. All patients underwent postoperative breast irradiation. The study demonstrated the specific timing with ICG diffusion throughout tumors after 3.3 ± 0.9 min before surgical excision. Total operating time averaged 192.9 ± 28.5 min, including sentinel lymph node biopsy or axillary lymph node dissection when indicated. The robotic lumpectomy component averaged 46.4 ± 10.5 min of console time. Notably, the surgical outcomes were successful with negative margins by frozen section examination in all 63 patients (100%). Analysis of ICG-guided surgical margins by permanent sectioning showed negative pathological margins in 62 of 63 specimens (98.4%). This surgical study represents one of the first demonstrations of a novel surgical technique with potential to reduce reoperation rates. The technique offers particular advantages for patients with larger, deeper breast tumors, potentially achieving results without visible breast scarring. Although the initial findings from this single institution are promising, multi-institutional studies are needed both the reproducibility of the technique and its long-term oncological outcomes.