Method Profile
Preprint combining human FELINE trial profiles, expression-based drug prioritization and selected fourteen-day drug tests in four human xenograft-derived organoid models.
Nominate and experimentally compare vulnerabilities of endocrine-insensitive epithelial populations in ER-positive breast cancer.
Nine human FELINE trial patients and four human patient-derived xenograft-derived organoid (PDxO) models: HCI003, HCI017, UIC013 and UIC020.
Clinical single-cell differential abundance, model-specific single-cell signatures, Drug Estimation from single-cell Expression Profiles (DREEP) predictions and organoid-area dose–response assays.
Human FELINE trial samples
Postmenopausal ER-positive/HER2-negative breast tumors sampled before and after fourteen days of letrozole; five responders and four nonresponders.
Human tumor-derived organoids
Human tumors passed through immunodeficient mice before organoid culture. Mouse-cell depletion/checks used; immune/inflammation context is reduced.
Computational drug reference
Drug Estimation from single-cell Expression Profiles (DREEP), using Cancer Therapeutics Response Portal version 2 cell-line pharmacogenomics for 450 computational compound scores.
Keep differential-abundance labels, computational drug scores and measured drug effects separate. Selected PDxO assays used image cytometry, two or three independent biological experiments depending on compound, and AUC comparison by one-way ANOVA with Dunnett testing.
Treat four PDxOs with endocrine drugs and selected DREEP-prioritized kinase inhibitors.
Four human PDxO models; two or three independent biological experiments per compound.
Vehicle and dose-response comparisons among models; fulvestrant/4-hydroxytamoxifen at 1 μM for fourteen days.
Partial endocrine growth responses persisted during long passage; EdU declined even where area differences were nonsignificant. UIC013 was more sensitive to dasatinib/pluripotin and to vandetanib than selected comparators; AZD8055 responses were similar across the four models.
Single-cell analysis of the FELINE breast cancer trial found diverse endocrine-insensitive populations. Four human tumor-derived xenograft organoid models supplied measured drug tests, with model-specific responses and limited patient-level longitudinal support.
