Photoacoustic imaging a PDT response marker for monitoring vasculature changes

1 Introduction

PAI is a technique depends upon the launch of acoustic waves from light-absorbing structures due to their thermal expansion. PAI is showing promises for several biomedical applications research in and translational research. Photodynamic Therapy (PDT) is a minimal invasive localized therapy that utilizes light-activated drugs called photosensitizers by initiating direct cell kill. PDT can be applied as a stand-alone therapy or as adjuvant therapy to present a standard of care. PDT is gaining popularity and exploits improvements in biomedical imaging techniques. Due to the higher sensitivity, shorter acquisition time and higher temporal resolution, they have achieved marvelous consideration in biomedical applications (Attia et al., 2019; Cao et al., 2017; Erfanzadeh and Zhu, 2019; Hosseinaee et al., 2020; Manwar et al., 2020; Zhao et al., 2019; Zhou and Jokerst, 2020). Tissues carrying phototoxic photosensitizers in molecular or nanoparticle form are irradiated with light of a particular wavelength for the generation of the excited singlet and triplet states resulting in reactive oxygen species (ROS), including singlet oxygen, superoxide anion, and other reactive species, such as the hydroxyl radical and hydrogen peroxide. The oxidative damage to various critical cellular components in the proximity of the ROS destroys the tumor due to either direct action on the tumor cells or necrosis induced by the disruption of the tumor's vasculature, cancer cell death, and destruction of unwanted tissue. Following systematic administration of a photosensitizing agent, it preferentially accumulates and is retained in the rapidly proliferating tumor tissues. Light is administered to the tumor site at times of maximum selective photosensitizer concentration in the tumor and the surrounding host tissue. In vivo studies of PDT disclose that the antitumor effect originates from different mechanisms, including direct damage to cancer cells and damage to vasculatures, impacting the Hemoglobin (Hb) levels.

The numerous PDT studies (Beard, 2011; Chen et al., 2021; Das et al., 2021; Dolet et al., 2021; Fu et al., 2019; Hester et al., 2020; Kim et al., 2015; Li and Chen, 2015; Liu and Li, 2020; Meng et al., 2019; Okumura et al., 2018; Qiu et al., 2021; Rao et al., 2020; Valluru and Willmann, 2016; Wang et al., 2016; Xu and Wang, 2006; Xu, 2011; Yu et al., 2017; Atif et al., 2016) indicated that this technology effectively treats a wide range of tumors by either direct cell kill or vasculature damage. The measurements of the tumor blood volume, blood flow, and Hb levels are essential for assessing tumor response in preclinical model systems and clinically.

Intravascular photoacoustic imaging is developing as a reliable image modality to assess tumors and post-treatment effects. Imaging contrast by endogenous or exogenous therapeutic agents increases the image modality's reliability and can deliver substantial benefits for response monitoring. The photoacoustic signal can be significantly improved through image contrast agents. Improvement in diagnostics can use therapeutics with photoacoustic imaging to treat tumors effectively. Deep tumor imaging, therapeutic responses monitoring, Hb detection, and ROS detection are important parameters for tumor growth and characterization. We utilized photoacoustic signals enhancement approaches on tumor models. Our study extracts PDT altered hemoglobin levels at different time intervals and correlated these results with tumor necrosis. The use of PAI and ultrasound has been previously proposed (Hester et al., 2020).

2 Materials and methods

Tumour were induced in the hind limbs of nude mice using PC3 cells. After the injection of PS, mice were imaged by PAI. ALA (60 mg/kg) and lower (1 mg/kg) & higher (3 mg/kg) concentration of BPD were used as photosensitizers. After administration of ALA and BPD for 4 hrs or 15 min respectively light activation was started. Tumours were imaged at multiple λ by the PAI system initially at 780 and 830 nm as the minimum order to determine the total hemoglobin signals to acquire the blood volume and oxygen level.

Early outcome indicators are suggested for this study as photodynamic therapy responses against tumors. Post Photodynamic therapy, Photoacustic (PA) images are retrived by different angles. Then images are superimposed on anatomical images as reported in the previous study (Atif et al., 2021). The photodynamic therapy response analysis is based on pre-PDT, immediate Post PDT i.e. 0 Hrs, then 4 Hrs, and 24 Hrs, respectively. These spatial domain figures are analyzed using thresholding and counting image voxels exceeding that threshold. A certain tumor area is analysed from the full image.

Region outside the tumor is compared with tumor area. Image is additionally disintegrated into constituent channels in the color scales, i.e., red, green, and blue. The spatial domain image is converted into an 8-bit greyscale.

Histograms and derivatives of these histograms for tumor area images are analysed for every case of Pre PDT, Post PDT, 4Hrs, and 24Hrs. Likewise, histograms and their derivatives for normal area images are calculated for each case of Pre PDT, Post PDT, 4Hrs, and 24Hrs. Spatial graphs are also transformed to the frequency domain to depict the frequency contents present in the tumor and normal tissues as detected by PAI.

This study was approved by the Princess Margaret Cancer Centre animal care committee, Toronto, Ontario, under AUP 6515.1.1.

3 Result and discussion

PAI uses pulsed or frequency domain near-infrared light mainly absorbed by specific marked component (e.g., hemoglobin), causing transient local heating (« 1 °C). It results in thermoelastic expansion of the target which leads to the emission of acoustic waves detected by an external ultrasound transducer. PAI delivers high resolution (50 µm) 3D absorption maps of vasculature, shown in Fig. 1 for BV.

Close

Fig. 1

An example of a subcutaneous prostate tumor treated with 5-ALA 60 mg/kg and 535 nm radiant exposure of 150 J/cm². The column shows a short-term increase in the BV followed by a strong decrease; however, there is no complete stasis.

Export to PPT

Fig. 1 shows a subcutaneous prostate tumor treated with 5-ALA 60 mg/kg and 535 nm radiation to a radiant exposure of 150 J/cm². The entire column of Pre-PDT, Post-PDT, 4hr PDT, 24 hr PDT shows a short-term increase in the BV followed by a substantial decrease but no complete stasis.

Table 1 shows the temporal evolution of Ultrasound and Photoacoustic imaging before and after PDT resulting in the spatial and frequency domain histograms of normal skin tissue layers and a subcutaneous tumor. Spatial intensity distribution histograms were prepared and evaluated for variations in the frequency domains for quantitative interpretation of experimental results.

Table 1 Original image of blood volume taken from Fig. 1, its gray scale version, PAI for the tumor region and normal skin for Pre PDT, Post PDT, 4Hrs and 24Hrs. An Example of the temporal evolution of US and PAI images prior to and post-PDT, the resulting spatial and frequency domain histograms of normal skin tissue layers, and a subcutaneous tumor. 1st-row full view containing US and PAI information, the 2nd-row highlight in center converted to greyscale, 3rd row, extracted PAI derived vascular volume component; 4th-row normal skin US and PAI data outside the PDT treatment field.

	Pre PDT	Post PDT	4 Hrs	24 Hrs
Original images
US Greyscale
PAI Tumour
Normal Skin

Fig. 2 shows the intensity differences between the tumor ROI and the skin ROI as a function of greyscale intensity. The US intensity ratio between the tissues is reduced over time. In particular, the high US intensities, representing large changes in the acoustic impedance mismatch, are disproportionally lost. Post PDT, an overall homogenization of the tissue is suggested through the treatment by 24 Hrs.

Close

Fig. 2

Ratio histogram of the tumor over normal skin Region of Interest (ROIs) from Pre PDT up to 24Hrs.

Export to PPT

An analysis of the US and the PAI images presented in Table 1 and the extracted intensity histograms and 2D spatial frequencies shown in Table 2 did not show statistically significant differences when approximated by a normal distribution. This holds true for the US and PAI data between the tumor and normal tissue, and the spatial frequency distributions' temporal evolution did not. No changes in the frequency distribution were noted for the tumor tissue intensities distributions, see Table 3, 2nd column. In contrast, for normal skin, the frequency mean shows a tendency to increase and the variance decrease, see Table 3 3rd column. The mean of the normal US spatial frequency distribution approximates that of the tumor, indicating an enhancement in the size of the structures causing acoustic impedance differences. The spatial frequency of the acoustic impedence changes can be explained by increased apoptosis in the normal skin tissue (Castano et al., 2006; Kessel, 2020; Plaetzer et al., 2005) whereas, given the high heterogeneity in the tumor tissue and the continuous presence of apoptotic cells therein. Modifications of the US response due to PDT-induced apoptosis was reported previously (Atif et al., 2022b; Czarnota et al., 1999; Li et al., 2006; Vlad et al., 2011; Wilson et al., 1999). For the PAI, the spatial frequency distributions' temporal evolution also does not show statistically significant differences for the tumor tissues. The number of observations for the normal skin tissue is low, and hence the variability in the mean of the frequency distribution shows more variability as a function of time appears more random. There is no apparent trend in the evolution of the center frequency or the width of the distribution. However, again the variability or change in the normal tissue is larger than in the tumor tissue. A possible source for this alteration between the tumor and normal skin response may be that the former has vasculature dominated by large, well-perfused vessels, whereas the latter's vasculature is dominated by arterioles, capillaries, and venules, providing a different response to PDT. While the use of PAI in PDT monitoring has received only limited attention for PDT, predominantly related to photosensitizer development (Li et al., 2017; Liu et al., 2021; Zhang et al., 2020), it is well established for monitoring vascular responses (May et al., 2016; Neuschmelting et al., 2018; Rich and Seshadri, 2015).

Table 2 Spatial and frequency Histograms of US encoded data as shown in Table 1.

	Pre PDT	Post PDT	4 Hrs	24 Hrs
Intensity Histograms Tumour
Intensity Histograms treated normal skin
Frequency HIstogram Tumor
Frequency Histograms treated normal skin
PAI intensity histogram Tumour
PAI Intensity Histogram Normal treated
PAI Frequency Histogram Tumour
PAI Frequency Histogram Normal Treated.

Where rate of change of oxyen is ratio of the difference of vales of oxygen levels from current to the previous with the change in time between the PDT observations.

In Fig. 3, the rate of change in blood volume is shown versus the rate of change for the three photosensitizer doses tested. The scatter plot of blood volume against oxygenation changes represents cells reduced metabolic activity, possibly including cell death.Where, rate of change of Oxygen is the difference between the current and previous values of Oxygen divided by the time difference between the observed values.

Close

Fig. 3

Rate of change of Blood Volume of ALA PDT and BPD low and a high dose of PDT versus rate of change of oxygen.

Export to PPT

In summary, structural alterrations of tissue were given less consideration, but the variations in acoustic impedance changes post-PDT was observed. The results represented a close spatial correlation between signal providing BV depending on the rate of change of oxygen. The experimental results from animals treated with ALA for Pre-PDT, Post-PDT, 4hr PDT, and 24 hr PDT show a short-term increase in the BV followed by a substantial decrease. However, there is no complete stasis based on these data. In another study, the ultrasound role and photoacoustic imaging were assessed to improve PDT-mediated outcome prediction for cancer. Photosensitizer tracking uptake and vascular obliteration were measured to study oxygen dynamics for overall estimation of tumor responses (Hester et al., 2020). In our group's earlier study, we discussed the quantification of blood volume, total hemoglobin saturation and ultrasound signal changes correlated with PDT-induced amendments in the tumor (Atif et al., 2022a). The present study applied photoacoustic signal enrichment methods to this tumor model. PDT-induced alterations of hemoglobin levels with different time intervals were explained, and results were correlated with tumor necrosis. Combined with structural imaging, these blood volume changes are an early indicator of a Photodynamic response.

4 Conclusion

In the current study, we studied the role of total hemoglobin and blood volume in the photodynamic activity and Ultrasound (US) signal changes and their correlation with PDT-induced alterations in the tumor. Images are represented in frequency domains by application of two-dimensional FFT on spatial domain images. Furthermore, variations in the acoustic impedance post-PDT have been demonstrated. Animal experimental results treated with ALA for Pre-PDT, Post-PDT, 4hr PDT, and 24 hr PDT show a short-term increase in the Blood Volume (BV) followed by a substantial decrease. However, there is no complete stasis. Photoacoustic signals augmentation approaches were implemented on tumor models and highlighted that PDT influenced variations of hemoglobin levels with different time intervals and correlated these outcomes with tumor necrosis.