Photoacoustic-guided photothermal therapy by mapping of tumor microvasculature and nanoparticle

1 Introduction

Cancer remains as a terrible disease with high morbidity and mortality worldwide [1]. Nowadays, immunotherapy [2], [3], [4], photodynamic therapy (PDT) [5], [6], [7] and photothermal therapy (PTT) [8], [9], [10], [11], [12], [13] have been proposed as precise and effective treatments for cancer. As an emerging therapy, PTT has made tremendous progress in improving the therapeutic outcomes, in which the photothermal agents utilize photoenergy to produce localized heat to induce the death of cancer cells. Different from conventional therapies, PTT has the following advantages: The laser wavelength and radiation dosage used can be flexibly switched to precisely kill the tumors and minimize damage to surrounding normal tissues; PTT is highly effective and noninvasive therapy that is capable of eliminating various types of cancers by using different biocompatible photothermal agents [14]. Although PTT has demonstrated its clinical value in fighting cancer, any treatment method needs the assistance of imaging technology to enhance treatment efficacy in clinical translation [15], [16], [17], [18]. Photoacoustic imaging (PAI) is a hybrid biomedical imaging technology that combines high-contrast optical imaging and deep-penetration acoustic imaging [13, 19–31]. Moreover, PAI and PTT are easily combined in biomedicine because the nanoplatforms used for PTT have the same physical properties as photoacoustic contrast agents, so the most nanoplatforms used for PTT can be used to enhance PAI at the same time. Recent years, PAI has gradually played a key role in improving cancer diagnosis and treatment, especially in precise pinpointing of tumors and monitoring of tumor response to therapy [18, 32–35]. As an important branch of PAI, photoacoustic microscopy (PAM) has great application prospects in PTT by observing the tumor microenvironment and the distribution of nanoplatforms with high resolution. However, due to the limitation of excitation laser and spatial resolution, currently there is no PAM system suitable for PTT, which can both reveal the tumor microvasculature and flexibly visualize the distribution of different nanoplatforms within the tumor. Therefore, there is an urgent need for a specific PAM system to precisely guide PTT and evaluate the therapeutic outcomes, especially in future clinical practice.

In this paper, we explored a wavelength-switchable PAM system, as a noninvasive and high-resolution imaging tool, to characterize the entire tumor vasculature and nanoparticles accumulation in tumor models for guiding PTT and evaluating its efficacy. Particularly, we aimed to investigate the ability of PAM to reveal the changes of tumor microenvironment, tumor volumes and the feeding arteries and draining veins around the tumor. Furthermore, the accumulation and distribution of Den-RGD/Cy7 nanoparticles via intravenous injection in 4T1 tumor model was monitored, the nanoparticles concentration at the tumor site reached its maximum at 8 h, which was used to determine the optimal irradiation timing. The comprehensive information of microvascular structures and nanoparticles can be used to guide the implementation of PTT. What’s more, the tumor volume and vascular morphology after treatment can be monitored and quantitatively analyzed to evaluate the therapeutic outcomes of PTT. The results proved that the proposed PAM can be a powerful tool for PTT.

2 Materials and methods

2.1 In vivo PA imaging system

Figure 1(A) shows the schematic diagram of wavelength-switchable PAM system (AOPM, Guangdong Photoacoustic Medical Technology Co.,Ltd). The pulsed laser (DTL-314QT, Laser-export), operating at wavelength of 532 nm with ∼7 ns duration, is used as excitation source for imaging microvasculature. First, the 532 nm laser beam passes through a spatial filter system (KT310/M, Thorlabs) to produce a Gaussian beam. Then, the 532 nm laser beam is focused by a 4× objective lens (NA = 0.1, Daheng Optics) to illuminate biological tissues. In addition, an optical parametric oscillator (OPO) laser (NT200, EKSPLA) outputs 650–2500 nm laser to visualize the distribution of nanoparticles. In this experiment, 750 nm laser is used to image the synthetic Den-RGD/Cy7 nanoparticles (Peak absorbance at 750 nm). Two wavelengths (532 and 750 nm) light are converged by a dichroic mirror and focused by the same 4× objective lens. A focused ringed ultrasonic transducer (8 mm O.D., 3 mm I.D., central frequency: 35 MHz) with focal length of 8 mm and −6 dB bandwidth of 120% is used to receive the PA signals. The piezoelectric element achieves self-focusing through a spherical structure. A 3D-printed tank filled with deionized water is used to couple the PA signals. The deionized water is in contact with the biological tissue through a transparent membrane. Finally, the PA signals generated by different lasers are used to reconstruct the microvasculature and nanoparticles in the same area. For imaging the tumor microvasculature, the step size of the X axis is 4 μm, and the step size of the Y axis is 10 μm. The laser pulse repetition rate is 10 kHz. For imaging the nanoparticles, the step size of the X axis is 16 μm, and the step size of the Y axis is 10 μm. The laser pulse repetition rate is 2.5 kHz. This scanning system takes about 10 min to image an area of 20 × 20 mm. By imaging nanoparticles, the distribution and concentration of nanoparticles can be obtained. The tumor volume can also be obtained by the distribution of nanoparticles in the tumor. The 532 nm laser can assess the condition in the tumor regions by visualizing the tumor microenvironment and the feeding vessels and draining vessels around the tumor with high resolution. In vivo PA imaging was used to assess tumor status and to determine the treatment timing and laser therapy dosage to optimize therapeutic efficacy (Figure 1(B)). After treatment, the microvascular changes can be investigated to evaluate the treatment effect of the tumor. Figure 1(C)–(F) show the morphology and distribution of microvasculature and nanoparticles (In situ injection, 0.5 mg/ml, 5 μl) in the mouse brain in different cross-sections and three-dimensional structures. The imaging results demonstrate that the proposed PAM has immense promise for imaging orthotopic tumors in multiple organs, especially in preclinical research. Furthermore, a sharp-edged surgical blade was imaged to characterize the resolution of the PAM system. The edge-spread function was calculated by taking the distribution of normalized PA signals amplitude along the blue dotted line in the inserted PA images. Taking the derivative yielded the line-spread function (LSF), Hence, the lateral resolution of the system as defined by the full-width at half-maximum (FWHM) of the LSF. Figure 2(A) indicates that the lateral resolution of 532 nm laser is 7.5 μm, and Figure 2(B) shows that the lateral resolution of 750 nm laser is 35 μm. Independent of the excitation laser, the axial resolution mainly depends on the center frequency and bandwidth of the transducer, the axial resolution of the developed dual-wavelength PAM system is about 42 μm, as shown in Figure 2(C).

Figure 1:

(A) The schematic diagram of the photoacoustic microscopy system. (B) Guiding diagram of photoacoustic imaging in photothermal therapy. (C–E) The merged photoacoustic images of microvasculature and nanoparticles distribution in a complete mouse skull state. (F) 3D photoacoustic images.

NDF, neutral density filter; L1, L2, lens; PH, pinhole; DM, dichroic mirror; OL, objective lens; UT, ultrasonic transducer.

Figure 2:

The imaging parameters of the PA system and the characterizations of the nanoparticles.

(A) The lateral resolution of the 532 nm laser imaging system. (B) The lateral resolution of the 750 nm laser imaging system. (C) The axial resolution of the 532 and 750 nm laser imaging system. (D) UV–vis spectra of Cy7 and Den-RGD/Cy7, respectively. (E) DLS data of the nanoparticles dispersions. (F) Photothermal effect of 0 mg/mL, 0.025 mg/mL, 0.1 mg/mL Den-RGD/Cy7 nanoparticles under irradiation with a 750 nm laser (1 W/cm²).

3 Results and discussion

The tumor microenvironment was investigated in vivo to demonstrate the ability of the PAM system to map the tumor microvasculature. Mouse of five weeks old were used in this experiment. The mouse was anesthetized by inhaling 1.5% isoflurane. Before 4T1 breast cancer cells subcutaneous injection (day 0), the mouse ear (Figure 3(A)) was first imaged with PAM system. Maximum amplitude projection (MAP) image (Figure 3(B)) and depth-encoded image (Figure 3(C)) can clearly reveal the arteries (A1, A2), veins (V1, V2) and microvasculature. After the control imaging, 4T1 breast cancer cells were injected into the imaging region of the mouse ear. A small tumor bulge was found at the injection site on day 4 (Figure 3(D)). PA images (Figure 3(E) and (F)) show more tumor angiogenesis at the tumor site, and the blood vessels around the tumor become more tortuous. The volume of the tumor became larger on day 7 (Figure 3(G)), and the blood vessels around the tumor also had a significant deformity. PA images (Figure 3(H) and (I)) show that the microvascular structure within the tumor was denser, and the diameter of the arteries and veins at the edge of the tumor also became larger. To further quantify the growth process of the tumor, the distributions of blood vessel diameter was counted on day 0, day 4, day 7 (Figure 3(J)). We found that tumor growth was accompanied by an abnormal increase in the number of microvessels. Then, the density of the blood vessels and the volume of the tumor were calculated separately (Figure 3(K)). The vessel density on day 7 was more than twice that on day 0. The tumor volume reached approximately 40 mm² on day 7. Finally, the diameters of two feeding arteries (A1, A2) and two draining veins (V1, V2) are counted (Figure 3(L)). From the results, we found that the diameters of the draining veins doubled on day 4, and then vein 2 also had obvious increase on day 7, while the feeding arteries only have a slight increase in diameter. The reason should be that the tumor enlarged and squeezed the blood vessels, and the reflux resistance of the draining veins became greater. As a result, the intravascular pressure increased. Because the venous walls are thinner and less flexible than the arterial walls, they tend to dilate and become larger in diameter easily. This experiment demonstrated that PAM can visualize and accurately evaluate the tumor microvasculature and the feeding arteries and draining veins around the tumor.

Figure 3:

In vivo monitoring of the tumor growth by photoacoustic microscopy.

(A) Mouse ear before injecting 4T1 tumor cells (day 0). (B) The maximum amplitude projection microvasculature image of the mouse ear on day 0. The white dotted box is the region of interest. (C) The depth-coded microvasculature image of the mouse ear on day 0. (D) The mouse ear on day 4. (E) The maximum amplitude projection microvasculature image of the mouse ear on day 4. (F) The depth-coded microvasculature image of the mouse ear on day 4. (G) The mouse ear on day 7. (H) The maximum amplitude projection microvasculature image of the mouse ear on day 7. (I) The depth-coded microvasculature image of the mouse ear on day 7. (J) The distribution of vascular diameter on day 0, day 4, day 7. (J) The distribution of vascular diameter on day 0, day 4, day 7. (K) The vessel density and tumor volume on day 0, day 4, day 7. (L) The diameters of the arteries and veins on day 0, day 4, day 7.

To demonstrate the PAM can reveal the subtle microvascular responses to PTT. Mice inoculated with 4T1 breast cancer cells were divided into three groups for different treatments. The control group only injected the Den-RGD/Cy7 nanoparticles (0.5 mg/ml, 0.1 ml) in situ. In the under-treated group, the Den-RGD/Cy7 nanoparticles (0.5 mg/ml, 0.05 ml) were injected in situ and then illuminated with CW 750 nm laser for 10 min (1 W/cm²). In the over treated group, the Den-RGD/Cy7 nanoparticles (0.5 mg/ml, 0.1 ml) were injected in situ and then illuminated with CW 750 nm laser for 10 min (1 W/cm²). After treatment, imaging of the tumor regions was performed at 6 h, 24 h, 48 h, and 72 h to evaluate the effect of PTT by monitoring structural changes in tumor vasculature and the surrounding tissue. Figure 4(A) shows the morphologic growth of the tumor in the control group. Figure 4(B) presents that the central region of the tumor in the undertreated group was damaged at 6 and 24 h, but the marginal region of the tumor was still abnormally expanded after 48 h. In the overtreatment group (Figure 4(C)), scabs appeared at the tumor site, and the vessels around the tumor were also damaged. When the scab fell off at 72 h, it was found that the entire tumor regions and surrounding tissues were damaged. To furtherly quantify the tumor microvascular in PTT, PA images were analyzed. The tumor volumes were analyzed in Figure 4(D). The tumor volume in the control group showed an upward trend. However, in the undertreatment group, the tumor volume first decreased briefly, and then began to increase. Finally, in the overtreatment group, the tumor became smaller, eventually scabs formed, and the tumor was eliminated completely. In Figure 4(E)–(G), the diameters of the feeding vessels and draining vessels marked by white arrows (Figure 4(E)), the amplitude of PA signals (Figure 4(F)) in tumor region and vascular tortuosity (Figure 4(G)) in white dotted rectangle also show the same changes as the tumor volume changes (Figure 4(D)). In Figure 4(E), the diameter of blood vessel (C1) increased at 72 h was caused by the exfoliation of scabs damaged the skin tissue, which led to more blood supply to the injured regions. Overall, these results demonstrate that photoacoustic microscopy provides a useful means to quantitatively depict tumor microenvironment and vascular morphology in PTT, which has immense promise to guide and improve therapeutic efficacy.

Figure 4:

Monitoring of tumors undergoing different in situ photothermal treatments.

(A) PA imaging results in the control group. (B) PA imaging results in the undertreatment group. (C) PA imaging results in the overtreatment group. (D) Tumor volumes in different experimental groups at different timing. (E) The diameter changes of the draining vessels in different treatment groups. (F) The amplitude of PA signals in tumor area. (G) Vascular tortuosity in white dotted rectangle. *: Due to the scab formation and exfoliation.

To verify the ability of PAM to characterize nanoparticle aggregation, a breast cancer model was imaged after injecting Den-RGD/Cy7 nanoparticles through the tail vein. Figure 5(A) shows the photograph of the tumor model, and then we imaging the tumor site using 532 nm laser to visualize the microvascular system of the tumor and surrounding tissues (Figure 5(B)). The imaging results showed that the tumor site was accompanied by an abnormal increase in blood vessels. Further, the distribution of nanoparticles and the resultant changes of PA signals inside tumor were continuously monitored (Figure 5(C)–(H)). Before Den-RGD/Cy7 injection through the tail vein, no obvious background PA signal was detected, owing to the low absorption of living tissues near 750 nm. After intravenous injection of Den-RGD/Cy7 at a concentration (0.5 mg/ml, 0.15 ml) into mouse, moderate PA signals were detected at the tumor site, which gradually increased over time and reached its maximum at 8 h postinjection (Figure 5(J)), suggesting the accumulation of Den-RGD/Cy7 nanoparticles in the tumor. Therefore, laser irradiation of tumor model was conducted at 8 h postinjection of nanoparticles. In order to study the relationship between the PA signals amplitude and the concentration of nanoparticles, the in vitro imaging was performed on samples with different concentrations. A good linearity was observed between the intensity of PA signal and the dose of Den-RGD/Cy7 nanoparticles, which indicates that Den-RGD/Cy7 has good potential for in vivo PAI and PTT.

Figure 5:

In vivo monitoring of tumors undergoing different photothermal treatments.

(A) The photograph of 4T1 tumor model. (B) Microvascular morphology of the tumor region was imaged with 532 nm laser. (C)–(H) In vivo non-invasive 750 nm laser PA imaging results of mice in 4T1 tumor model at different time points after Den-RGD/Cy7 nanoparticle via tail vein injection. (I) The imaging results and the PA signals statistics of nanoparticles at different concentrations under 750 nm laser. (J) PA signals amplitude of nanoparticle in the tumor center.

Finally, the tumor microvasculature was monitored to evaluate the active targeted therapy efficacy. The Den-RGD/Cy7 nanoparticles were injected through the tail vein. At 8 h postinjection, the tumor site was treated with CW 750 nm laser for 10 min (1 W/cm²). PA imaging was performed at the tumor site before and at 6, 24, 48, and 72 h after nanoparticles injection. As can be seen from the PA images (Figure 6(A)–(G)), the tumor volume gradually decreases after PTT kills the tumor cells. It is worth noting that edema around the tumor site appeared 24 h after laser irradiation due to tissue reflection. At 72 h (Figure 6(E)), the edema around the tumor completely disappeared, blood vessels were clearly shown in the normal tissues around the tumor, and the tumor site was seriously damaged. The imaging results illustrate that PTT is a very promising way to fight cancer with minimal invasiveness and high effectiveness. The results show that the PTT has achieved better treatment efficacy. Figure 6(F)–(G) are the cross-sectional maximum amplitude projection images in tumor site at 72 h. It can be seen from the images that there is no obvious vascular morphology inside the damaged tumor after PTT, and the normal tissues around the tumor have not been damaged. Figure 6(H) depicts that the tumor volume continues to decrease after PTT, and the tumor volume measured is about 55 mm³ at 72 h. The amplitude of PA signals in the center of the tumor was stronger at 6 h than before the intravenous injection, because of the enhancement of the PA signals caused by the large accumulation of Den-RGD/Cy7 nanoparticles in the tumor. The PA signals amplitude in the tumor site decreased with tumor damage and longer circulation time. Furthermore, the hematoxylin and eosin (H&E) staining analyses were conducted on R1 and R2 marked in Figure 6(E). The normal tissue (R1) around the tumor did not show obvious cell necrosis (Figure 6(I)). The tumor site (R2) exhibited severe tumor tissue damage with high percentage of cell necrosis (Figure 6(J)). The results of H&E staining were consistent with those of PA images, indicating PAM has ability to evaluate the therapeutic effects of PTT noninvasively.

Figure 6:

In vivo photothermal therapy after injection through the tail vein.

(A)–(E) The microvasculature was imaged at different timing to evaluate the therapeutic effect. (F) and (G) The tomographic projection images of the tumor region at 72 h after treatment. (H) Tumor volumes and PA signals amplitude at different timing after treatment. (I) Image with H&E staining histopathology of normal tissue around the tumor (R1) in Figure E. (J) Image with H&E staining histopathology in tumor site (R2) in Figure E.

4 Conclusion

In conclusion, a wavelength-switchable PAM was proposed to guide and assess the PPT by mapping the microvascular morphology and the distribution of nanoparticles with high resolution. The PAM system can reveal morphological changes in the tumor microenvironment and feeding arteries and draining veins following tumor growth. The PAM can also flexibly select near-infrared laser according to the absorption of the nanoplatforms to visualize the nanoparticles aggregation within tumor, which was used to determine the optimal irradiation timing. The in vivo experiments demonstrate the proposed photoacoustic microscopy enables the detailed quantitative visualization of tumor microvasculature and nanoparticles accumulation to guide PTT and noninvasively evaluate therapeutic effect, which may enhance PA imaging and improve therapeutic efficacy, and can be extended to guide and assess a larger field in precision medicine.