Background: A subset of natural killer cells lacking FcεRIγ expression (g-NK cells) demonstrate enhanced expansion, cytokine production, antibody-dependent cytotoxicity, and enhanced efficacy of therapeutic antibodies in preclinical models (Bigley, Blood Advances 2021). The monoclonal antibody, trastuzumab, is standard of care for treatment of HER2+ breast cancer in adjuvant and metastatic settings. We sought to determine whether the abundance of g-NK cells is associated with improved outcomes in trastuzumab-treated patients. Methods: We identified g-NK cells in single-cell (scRNA-seq) datasets from melanoma (n=47), renal (n=8), breast (n=29), and lung cancer (n=36) samples using predefined markers and defined a g-NK gene signature. We split the scRNA-seq cohorts into training and test sets and estimated cell-type compositions in pseudo bulk using MuSiC bulk tissue deconvolution (Wang, Nature Communications, 2019). We deconvoluted g-NK cells in 8 bulk RNA-seq HER2+ breast cancer cohorts (Table 1) and used Cox proportional hazard models to assess clinical outcomes stratified by g-NK and trastuzumab status. We reproduced our findings using real-world cohorts from patients sequenced via the Tempus xT test (Table 1). Results: The Spearman’s correlation coefficient between true and estimated pseudo bulk proportion of g-NK cells was 0.71 (p≤0.05). In a meta-analysis, the abundance of g-NK cells was associated with a lower hazard ratio (HR) for overall survival (OS) in trastuzumab-treated patients (cohorts 1-4; HR = 0.61; 95% CI, 0.37-0.99; p=0.04) but not trastuzumab-naive patients (cohorts 5-8; HR = 0.96; 95% CI, 0.80-1.15; p=0.65). In real-world, HER2+ breast cancer cohorts treated with trastuzumab, higher real-world progression-free survival (rwPFS) was associated with abundance of g-NK cells (N=104, HR = 0.78; 95% CI, 0.64-0.96; p=0.02). A similar trend was observed in real-world trastuzumab-treated gastric cancer patients (N=79, HR = 0.85; 95% CI, 0.72-1.00; p=0.06). Conclusions: G-NK cells are associated with improved OS and PFS in trastuzumab-treated patients from both clinical trials and real-world datasets, suggesting their presence may serve as a predictive marker for treatment efficacy. Adoptive transfer of g-NK cells is being explored to augment the efficacy of therapeutic monoclonal antibodies.

Datasets used in analyses: CALGB 40601 trial, C1-2 and C5; Finland Herceptin trial, C3-4 and C6-7; Molecular Taxonomy of Breast Cancer International Consortium, C8; Tempus Labs, RWD1-2. Abbreviations: NAT, neoadjuvant therapy; AT, adjuvant therapy; TB, tumor bank; RWD, real-world data.