Supramolecular attack particles (SPAPs) are ~110 nm extracellular glycoprotein particles released from cytotoxic T cells and natural killer cells that can autonomously kill cancer cells. The goal of the PoC was to produce SMAPs in sufficient quantity and purity to use in preclinical cancer models in mice and initiate preclinical testing in models such as glioblastoma or ovarian cancer. We used a natural killer cell line, NK-92, as a source of SMAPs. We compared affinity isolation and size exclusion chromatography to enrich SMAPs. The initial material from NK-92 cells grown in bioreactors was 90% extracellular vesicles and 10% SMAPs. The affinity based methods to remove extracellular vesicles was not efficient and could only remove 50% of the contaminating particles. An alternative approach based on three steps- tangential flow filtration, Fast Flow Superose 4B size exclusion chromatography (SEC) and Sephacryl S-1000 SEC was more successful in generating fractions that were 90% extracellular vesicles (fraction 2) and 90% SMAPs (fraction 3). Importantly, proteomic analysis revealed that survival promoting serpins were enriched in fraction 2 whereas cytotoxic proteins were enriched in fraction 3. Based on these findings we initiated experiments in mice with PANC1 cells and B16 melanoma models to look at toxicity and efficacy.