Chemotherapeutic effects on breast tumor hemodynamics revealed by eigenspectra multispectral optoacoustic tomography (eMSOT)

Abstract

Non-invasive monitoring of hemodynamic tumor responses to chemotherapy could provide unique insights into the development of therapeutic resistance and inform therapeutic decision-making in the clinic. Methods: Here, we examined the longitudinal and dynamic effects of the common chemotherapeutic drug Taxotere on breast tumor (KPL-4) blood volume and oxygen saturation using eigenspectra multispectral optoacoustic tomography (eMSOT) imaging over a period of 41 days. Tumor vascular function was assessed by dynamic oxygen-enhanced eMSOT (OE-eMSOT). The obtained in vivo optoacoustic data were thoroughly validated by ex vivo cryoimaging and immunohistochemical staining against markers of vascularity and hypoxia. Results: We provide the first preclinical evidence that prolonged treatment with Taxotere causes a significant drop in mean whole tumor oxygenation. Furthermore, application of OE-eMSOT showed a diminished vascular response in Taxotere-treated tumors and revealed the presence of static blood pools, indicating increased vascular permeability. Conclusion: Our work has important translational implications and supports the feasibility of eMSOT imaging for non-invasive assessment of tumor microenvironmental responses to chemotherapy.

Keywords: Taxotere; breast cancer; chemotherapy; docetaxel; optoacoustic; photoacoustic.

Figure 1

Taxotere inhibits the growth of KPL-4 breast tumor xenografts. (A) Schematic illustrating MSOT imaging and treatment scheduling. KPL-4 tumor bearing mice (n = 7 per treatment group) were injected on d22 post implantation with vehicle (saline) or a high priming dose of Taxotere (25 mg/kg, i.p.) (red arrow), after which vehicle or Taxotere (10 mg/kg, i.p.) was administered at the indicated time points (blue arrows). (B) Tumor growth curves for mice treated with saline (black line) or Taxotere (blue line). (C) Weight changes for saline (black line) and Taxotere (blue line) treated KPL-4 tumor-bearing mice. (D) Photographs of KPL-4 tumors of sacrificed mice obtained at the end of the imaging experiment (day 63 post implantation). Values in graphical plots are reported as mean ± SD (n = 7 mice per treatment group). *2 vehicle-treated mice were sacrificed at earlier time points (d36 and d42 post implantation) due to development of tumor ulceration (see Methods).

Figure 2

Spatiotemporal changes in oxygenation and blood volume of KPL-4 tumors along the course of Taxotere chemotherapy assessed by MSOT/eMSOT. (A) Mean whole tumor total hemoglobin concentration (THb) and (B) eMSOT oxygen saturation (sO₂) values obtained from KPL-4 tumors of mice treated with either vehicle (saline, black lines) or Taxotere (blue lines) for days 0-41 post treatment. During the course of Taxotere therapy the average tumor sO₂ gradually decreased, while an increase in THb was observed on the last day of the imaging experiment (d41). (C-D) Longitudinal changes in blood volume and oxygenation of a single representative KPL-4 tumor from vehicle (C) and Taxotere (D) treatment groups visualized by serial MSOT imaging at 7 consecutive time points. Middle panels: central optoacoustic cross-sections showing THb (800 nm, isosbestic point) distribution. Upper panels: THb maximum intensity projections (THb-MIP). Lower panels: pseudo-colorized eMSOT maps of tumor oxygen saturation overlaid on corresponding anatomical images. Color scale bars at the bottom of eMSOT images represent sO₂ levels ranging from 0% (green) to 100% (red). Data are displayed as mean ± SD (n = 7 mice per treatment group). *p < 0.05, **p < 0.01. Statistical significance between the two treatment groups was assessed by an unpaired two-tailed t-test. Scale bar; 2 mm.

Figure 3

Oxygen-enhanced eMSOT (OE-eMSOT) reveals altered oxygenation response in Taxotere-treated KPL-4 tumors subjected to an oxygen challenge. (A) Average kinetic curves showing mean whole tumor sO₂ values at each time point of the time course in vehicle (saline, black line) and Taxotere (blue line) treated mice subjected to an oxygen challenge (n = 5 tumors per group). The thick grey and blue sigmoidal curves show nonlinear fit of data. Taxotere-treated KPL-4 tumors exhibit little enhancement in whole tumor sO₂ during the oxygen challenge, indicating diminished vascular function. (B) Average kinetic curves showing mean tumor rim (saline; black line, Taxotere; blue line) and core (saline; black dotted line, Taxotere; blue dotted line) sO₂ values at each time point of the time course. The thick grey and blue sigmoidal curves show nonlinear fits of data. (C-D) eMSOT images of a single central optoacoustic tumor cross-sections from representative vehicle (3 left panels) and Taxotere (5 right panels) treated mice acquired during medical air (21% O₂) (C) and pure oxygen (100% O₂) (D) breathing conditions. Bottom panels: The presence of hemorrhagic blood pools in Taxotere-treated tumors #1 and #3 is indicated by yellow arrows. Color scale bars on the right of eMSOT images represent sO₂ levels ranging from 0% (green) to 100% (red). The switch from 21% to 100% O₂ is denoted by a vertical green line. Scale bars; 2 mm.

Figure 4

Histopathologic assessment of tumor hypoxia and vascularity confirms in vivo optoacoustic data. (A-F) Micrographs of central histologic whole tumor cross sections and corresponding optoacoustic slices from representative vehicle (left panels) and Taxotere-treated (right panels) tumors; (A) Cryosection block color photography. (B) THb distribution (800 nm, isosbestic point). (C) CD31+ microvessel distribution. (D) DAPI nuclear staining (blue) merged with CD31 (yellow) immunofluorescence signal. (E) eMSOT mapping of tumor oxygen saturation (sO₂) overlaid on corresponding anatomical images. Color scale bars on the left of eMSOT images indicate sO₂ levels ranging from 0% (green) to 100% (red). (F) DAPI (blue) merged with pimonidazole (green) immunofluorescence. Insets 1 and 2 in (D) and (F) represent successively magnified regions enclosed in the corresponding white-dotted rectangles. (G) Pearson's correlation analysis of optoacoustic and histopathologic data sets showing inverse correlation (R² = 0.582) between pimonidazole-positive hypoxic fraction (HF) and whole tumor eMSOT-sO₂ estimates. Correlations were assessed on a per-tumor basis from pooled HF/eMSOT-sO₂ measurements from vehicle (black dots) and Taxotere-treated (blue dots) mice (see Materials and Methods). (H) Tumor hypoxic fraction (HF). (I) CD31+ microvessel density (MVD). (J) Tumor necrotic fraction (NF). Data are displayed as mean ± SD (n = 5 mice for vehicle and n = 7 for Taxotere-treated groups respectively). **p < 0.01, ***p < 0.001. Statistical significance between the two treatment groups was assessed by an unpaired two-tailed t-test. Scale bar; 2 mm.

Figure 5

Taxotere induces the formation of blood lakes in KPL-4 tumors. Co-registered histologic and optoacoustic image pairs of two distinct transversal planes obtained from a single KPL-4 tumor at the end of Taxotere treatment. The two cross-sectional tumor planes (denoted by #1 and #2) are spaced by 1 mm. (A) Cryosection block color photography. The presence of a blood lake in tumor cryosections is indicated by a yellow arrow. (B) MSOT images showing THb distribution (800 nm, isosbestic point). (C) CD31+ immunofluorescence (yellow) merged with DAPI fluorescence (blue). Insets are magnified views of the corresponding regions enclosed in the white-dotted rectangles. (D) eMSOT mapping of tumor oxygen saturation (sO₂) shows that the blood lakes in #1 and #2 are filled with de-oxygenated blood. Color scale bars on the left of eMSOT images indicate sO₂ levels ranging from 0% (green) to 100% (red). (E) pimonidazole immunofluorescence (green) overlaid with DAPI (blue). (F) CA9 immunofluorescence (red) overlaid with DAPI (blue). (G) Magnified insets of the corresponding white-dotted rectangles in (E) and (F) showing merged DAPI, pimonidazole and CA9 immunofluorescence.