exosome secretions from colorectal cancer samples
Owner
Therapeutic Relevance
The mechanism is scientifically plausible and targets a well-established clinical problem: the high rate of unnecessary colonoscopies following positive FIT results (~79% find no advanced neoplasm). The EV proteomic triage approach is grounded in sound biology — tumor-derived extracellular vesicles carrying cancer-associated cargo proteins detectable in plasma. The proposal targets a biologically relevant disease mechanism (CRC biomarker shedding via EVs) and is strategically positioned in the post-FIT enriched population (21% advanced neoplasia prevalence) rather than average-risk screening, which is a smart mechanistic and statistical choice. However, the score is capped at 4 rather than 5 because: (1) the evidence base for EV proteomics in CRC is thin (median N=100 across 56 studies, only 15/51 CRC studies with independent validation), (2) only 14 proteins replicate across ≥2 studies and a third may be co-isolation artifacts, (3) the team candidly acknowledges 'the biology is the bet, and the evidence for it is thin,' and (4) no experimental data have been generated yet — all claims rest on literature meta-analysis. The plausibility is strong but unproven for this specific application.
Therapeutic Optionality
The concept has moderate flexibility. The EV proteomic platform and DIA-MS workflow could theoretically be extended to other cancer types where liquid biopsy triage is needed (e.g., lung, pancreatic, ovarian cancers following initial screening signals). The capture-tractable targets (EPCAM, EGFR, MET, CD44) are broadly expressed across epithelial cancers, suggesting some cross-cancer applicability. The pre-analytical SOP and analytical pipeline being developed would be transferable. However, the proposal is tightly focused on a single clinical use case (FIT triage for CRC), and the candidate protein watchlist is CRC-specific. The concept is not inherently a platform play — it is a single-indication diagnostic test. The pre-specified failure branches show some adaptability (e.g., repositioning as a reflex test, pivoting to single-marker design), but these are contingencies within the same indication rather than expansion to new therapeutic areas. Score of 3 reflects reasonable but not exceptional optionality.
Intellectual Property
The concept demonstrates strong novelty for IP purposes. No registered trial evaluates an EV proteomic panel for FIT triage at any scale (out of 116 CRC trials reviewed). The competitive landscape analysis shows all major pivotal programs (Freenome, Guardant, Exact Sciences, Geneoscopy) target average-risk screening with different modalities (ctDNA, multi-target stool DNA, stool RNA, multiomics), not post-FIT triage with EV proteomics. The closest analogue is a single n=400 trial at City of Hope for early-onset CRC detection. Key patentable elements include: (1) the specific EV proteomic panel composition (≤12 proteins + 2 normalisers) once locked, (2) the FIT-triage clinical positioning, (3) the combination of immunocapture ligands (EPCAM/EGFR/MET/CD44 ectodomains) with quantified analytes, and (4) the integrated decision threshold. The score is 4 rather than 5 because: the individual protein targets are known from published literature (14 candidates from existing studies), the EV isolation methods (SEC, UC, immunoaffinity) are established, and the absence of competitors could indicate the field has evaluated and deprioritized this approach rather than an untapped opportunity — a risk the proposal itself acknowledges.
