1093 Background: Circulating tumor cells (CTCs) expressing epithelial markers (EPCAM, cytokeratin (CK)) and lacking CD45 (a leukocyte marker) have been associated with poor outcome in many cancer types. Nonetheless, the presence of cells expressing both CK and CD45 (CK+/CD45+), circulating in the blood of cancer patients (pts) have also been reported, but not widely investigated. Early evidence indicates that circulating dual-positive cells (DPcells) are hybrids deriving from the fusion of tumor cells and macrophages. We previously reported that it is possible to detect DPcells in the blood of pts with metastatic breast cancer (BC) and that they are associated with shorter progression-free survival (PFS), in pts with <5 CK+/CD45- CTCs. Here, we investigated the impact of DPcells on overall survival (OS) in pts with advanced BC (aBC). Methods: Blood samples (7.5 ml) were collected from aBC pts before starting a new therapy and processed with the FDA-approved CellSearch platform for CTCs and DPcells enumeration. The prognostic role of CTCs and DPcells was assessed through the Kaplan-Meier method using the log-rank test. Single DPcells were isolated using the DEPArray platform and underwent whole genome amplification and lowpass whole genome sequencing (Ampli1 WGA and Ampli1 Lowpass kits). Results: Blood samples from 341 pts with luminal (n=168), HER2+ (n=76) and triple negative (n=88) BC were analyzed. Of these, 131 samples (38.4%) contained ≥5 CTCs (CTCpos), whereas DPcell were detected in 152 samples (44.6%, range 0-53), of which 66 (43.4%) were CTCpos and 86 (56.6%) CTCneg. Overall, DPcells were associated with a shorter OS: median OS 24.5 vs 35.0 months, p=0.046. However, when analyzing CTCpos and CTCneg separately, only the latter group showed a difference in OS according to DPcells presence. In particular, among CTCneg pts, those with ≥4 DPcells showed a 2.3-fold shorter OS (26.7 vs 60.6 months, p=0.025). Moreover, pts with ≥4 DPcells were less likely to experience a 6-months PFS clinical benefit (p=0.015). Interestingly, in the analysis by BC subtype, DPcells were confirmed to be associated with worse OS only in pts with triple negative BC (median OS 11.5 vs 16.9, p=0.048). To explore the exiology of DPcells, 2 out of 3 cells analyzed after single-cell isolation from 1 patient were confirmed to have copy number alterations (CNA) consistent with malignant cells. CNA and mutational profiling of additional single DPcells and CTCs are ongoing. Conclusions: DPcells are associated with worse OS in aBC pts, with the prognostic impact primarily in pts with <5 CTCs and triple negative BC. This suggests that DPcells might be an alternative way of tumor dissemination in specific pts, in which CK+/CD45- CTCs are less prevalent. More studies are needed to better elucidate DPcell clinical significance in BC, and to confirm their fusion-hybrid origin.