“…Although this model now appears to be both overly simplistic and fundamentally flawed, it did provide the basis for the development of rapid (e.g., 48 h incubation with a test drug), high throughput multiwell plate colorimetric (e.g., tetrazolium-based; crystal violet-based) and fluorometric (resazurin-based, such as CellTiter-Glo) assays for the assessment of cancer cell death post-treatment. These and other widely used preclinical assays (including colony forming ability, immunoblotting, flow cytometry, tumor growth delay in live animals) are not designed to recapitulate the degree of cellular and molecular complexity and heterogeneity that exists within a single tumor (intratumor heterogeneity) (reviewed in, e.g., [ 4 , 91 , 92 , 93 , 94 , 95 ]; also see Figure 2 B). Furthermore, continuous cell treatment with relatively high concentrations of chemotherapeutic drugs, which is required in order to induce a significant inhibitory effect (e.g., 50% “cytotoxicity”) in multiwell plate assays (see, e.g., [ 96 , 97 , 98 , 99 , 100 , 101 , 102 , 103 ]), often has little clinical relevance because many drugs are administered to a patient as a bolus [ 18 , 19 ].…”