{"entity":{"id":"ex-vivo-tumour-fragments","kind":"technology","name":"Ex vivo tumour fragments and slices","aka":[],"tldr":"Living tumour pieces can retain local cells and architecture for a defined experiment outside the body, while losing circulating supply and recruitment.","summary":"Fragment and slice preparations retain selected native tissue relationships rather than expanding a selected culture. Composition and function change with sampling, cutting, diffusion, medium and time. A retained vessel does not demonstrate perfusion, and visible immune cells do not establish the required interaction. Ladan's breast-slice assay and Salmon's human lung-slice migration experiments support different questions; neither validates every fragment test or treatment decision.","status":"emerging","asOf":"2026-11-13","links":[{"label":"Ladan et al. 2023 primary slice assay","url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC10542784/"},{"label":"Salmon et al. 2012 primary lung-tissue interaction","url":"https://doi.org/10.1172/jci45817"},{"label":"Hafner et al. 2016: readout interpretation, not fragment validation","url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC4887336/"},{"label":"FDA: statistical guidance for diagnostic test studies (March 2007)","url":"https://www.fda.gov/regulatory-information/search-fda-guidance-documents/statistical-guidance-reporting-results-studies-evaluating-diagnostic-tests-guidance-industry-and-fda"}],"tags":[],"related":[],"cancers":[],"sections":["drug-discovery"],"technologies":["functional-drug-testing"],"targets":[],"drugs":[],"companies":[],"institutions":[],"pathways":[],"terms":[],"trials":[],"people":[],"bottlenecks":[],"keyPapers":[],"journals":[],"dependsOn":[],"notes":["Preparation and assay card. Input: suitable viable tissue with donor/site/time, collection and transport conditions, dimensions and culture window. Cutting, quality checks and endpoints consume finite material and limit available conditions. Compare baseline quality and untreated or carrier controls with defined exposures. Attribute counts, stain fractions, movement or metabolic signal to the relevant cells, with units, time and denominator. Thresholds are method-specific rather than a universal fragment score. Fragment names the preparation; a drug screen names an experiment it may support. Different pieces from one tumour can contain different populations, so matched treatments still need composition checks.","Primary preparation example. Ladan used 300 micrometre slices and three-day exposure in a defined anthracycline-based assay, measuring morphology, proliferation and apoptosis. Diffusion into slices differs from vascular delivery; host metabolic activation, clearance, organ toxicity and new immune recruitment were not recreated. Preparation injury, poor viability, uneven diffusion, changing populations and non-tumour signal can invalidate interpretation. A failed culture is not evidence of clinical resistance. Contamination is another reason for assay failure; short-term response does not establish durable elimination, systemic safety or benefit.","Endpoint and validation example. Ladan reported 23 usable assays among 30 tumour-containing biopsies; three of the 23 did not proceed to the clinical treatment comparison, leaving 20 evaluable patients and 15 concordant MRI comparisons. The agreement compared sensitive, intermediate and resistant categories after four courses of anthracycline-containing chemotherapy; it is not a binary drug-killing rate. The study did not observe an ex vivo resistant group, limiting assessment of resistance detection. Those denominators, assay version and preoperative setting cannot validate another postoperative choice. Salmon measured matrix-associated T-cell location and movement in viable lung-tumour slices, not treatment-selection benefit.","Fictional reading exercise. Two fragments with different starting malignant-cell fractions both lose total metabolic signal. Equal tumour killing has not been shown. Verify composition and an attributed death endpoint against controls. Retained neighbours can be useful for an immune question only when their viability and relevant function are demonstrated.","Ladan's real study began with 41 consenting people: three did not undergo biopsy and eight biopsies lacked tumour cells, leaving 30 adequate tumour biopsies. Six assays failed the prespecified baseline-apoptosis criterion and one was lost to infection. Of 23 usable results, three people did not receive the chemotherapy, leaving 20 clinical comparisons. The main-text counts are used here; Table 2's footnote instead mentions two people not receiving chemotherapy. The reported interval for 15/20 agreement was 53 to 89 percent, and the pilot's prespecified target was at least 55 percent. Reaching that target supports further investigation, not a demonstration that using the assay to choose care improves outcomes.","The reported breast-study totals imply two sensitive dish calls with sensitive MRI, three sensitive dish calls with intermediate MRI, two intermediate dish calls with sensitive MRI and thirteen intermediate calls matching MRI. Thus 15/20 agree, including thirteen matches in the common intermediate category. Two of four MRI-sensitive tumours were called sensitive in the dish (50 percent), and two of five sensitive dish calls matched MRI sensitivity (40 percent). These are stated comparisons with an imperfect MRI outcome, not validated sensitivity or positive predictive value against known truth. FDA's March 2007 diagnostic-test guidance distinguishes agreement from accuracy when the comparator is not a reference standard. The study's sensitive category required more than 50 percent size reduction after four chemotherapy courses; it is not simply the standard RECIST partial-response category. Usable coverage is 23/30, about 77 percent; a matching clinical comparison was available for 15/30, or 50 percent, including failures and the untreated people in that denominator. Neither ratio measures postoperative benefit."],"principle":"Maintain viable tissue pieces for a defined comparison while measuring retained populations, quality and an attributed response.","strengths":[],"limitations":[]},"route":"/technologies/ex-vivo-tumour-fragments/","neighbours":{"section":[{"id":"drug-discovery","kind":"section","name":"Drug Discovery Platforms","route":"/fronts/drug-discovery/"}],"technology":[{"id":"functional-drug-testing","kind":"technology","name":"Functional (ex vivo) drug testing","route":"/technologies/functional-drug-testing/"}]}}