Highly dynamic mitotic-spindle microtubules are among the most successful targets for anticancer therapy. Protein-bound paclitaxel (Abraxane), one of the most successful microtubule-targeted chemotherapeutic drugs, has been reported to suppress the spindle-microtubule dynamics in tumor cells which results in the slowing or blocking of mitosis at the metaphase-anaphase transition and induction of apoptosis. In the current study, we hypothesized that temporal diffusion spectroscopy, which measures the variation of ADC values over a broad range of effective diffusion times using a combination of PGSE and OGSE acquisitions1, can quantify tumor microstructural changes in response to microtubule-targeted chemotherapy. This hypothesis is elucidated and validated using both well-characterized cell culture and in vivo MR experiments.Theory: Temporal diffusion spectra are acquired using oscillating diffusion gradients to replace conventional bipolar gradients, and short effective diffusion times can be readily achieved with only moderate frequencies. More importantly, by tuning oscillating frequencies, temporal diffusion spectra can be obtained to reveal microstructural information from short (intracellular) to long (subcellular) length scales. In vivo experiments: Two types of human ovarian cancer cells, OVCAR-8 (responder) and NCI/ADR-RES (resistant), were used to create tumors in mice. Animals were treated i.p. with either Abraxane or PBS (drug vehicle) for 4 days (10 mg/kg, every other day). Diffusion-weighted images (DWI) were collected before, and after 4-day treatments. The apparent diffusion spectra over frequencies (f) from 50-350 Hz were obtained with two b-values (0 and 400 sec/mm2), while the ADC at 0 Hz was estimated by a conventional pulsed gradient sequence with diffusion time = 48 ms. Tumor tissues were collected for histological analysis of caspase-3 activation immediately after the second imaging time point. In vitro experiments: Cultured OVCAR-8 and NCI/ADR-RES cells were treated with Abraxane at three different concentrations (0, 100, and 200 nM) for 24 hours. Cell samples were then collected, fixed, transferred to a tube, and centrifuged to form cell pellets for DWI experiments. In addition, cell size were quantified using both flow cytometry and light microscopy. Data analysis: Tumors were modeled as tightly-packed spherical cells, and the entire ADC spectra of each tumor were fit to the equation: ADC(f) = ADCres(d, Dinf-D0 ) + D0, where ADCsphere arises from intracellular space with restricted size d, infinite-high-frequency diffusion coefficient Dinf, and zero-frequency diffusion coefficient D0. For tumor tissues, Dinf is dominated by intracellular diffusion coefficient and D0 is associated with the extracellular tortuosity.Figure 1 shows that the fitted mean restricted size of Abraxane-treated OVCAR-8 cells increase significantly (p<0.05) with the increasing drug concentration. By contrast, there is no significant changes in restricted size of Abraxane-treated NCI/ADR-RES cells. The increase in restricted size for Abraxane-treated OVCAR-8 cells reflects the increase of tumor cell size as a response to antimitotic-treatment, which is confirmed by i) changes in mean cell size measured by light microscopy (Figure 2) and ii) changes in the width of forward scatter signal (FSC-W) which is proportional to cell size (Figure 3). For the in vivo experiments, the efficacy of Abraxane in treating OVCAR-8 tumors is validated by calculating the percentage of caspase-3 activation area from the entire caspase-3 staining slides. As shown in Figure 4, the percentage of caspase-3 activation area for Abraxane-treated OVCAR-8 is significant higher (p<0.01) than that of other three groups. In Figure 5, 6 out of 7 Abraxane-treated OVCAR-8 tumors show an increase of fitted mean restricted size, consistent with our in vitro observations. Interestingly, there is one Abraxane-treated OVCAR-8 tumor showing a 25% decrease of mean restricted size. It is also noted that this tumor has the largest percentage of caspase-3 activation area (65%) in all the treated-responder tumors, suggesting that the cell shrinkage resulted from apoptosis overturn the increasing trend of cell size resulted from mitosis inhibition. There is no significant differences in the other fitted microstructural parameters (Dinf and D0) and PGSE-derived ADC values, suggesting these parameters did not change at early stages after treatments. Temporal diffusion spectroscopy detects antimitotic-therapy-induced microstructural variations in tumor tissues which occur before frank changes in tumor size (data not shown) and conventional PGSE-derived ADC values.