Photodynamic therapy (PDT) is a minimally invasive treatment for cancer and other diseases that uses light-activated compounds to attack malignant cells and tissues. Using a photosensitizer (an agent that transfers energy from incident light), PDT produces reactive oxygen species—predominantly singlet oxygen (1O2)—that are toxic to the targeted cells. Direct imaging of singlet oxygen near-IR (NIR) luminescence at around 1270nm reveals the spatial and temporal heterogeneity of tumors and their response to PDT. However, the imaging process is technically challenging because of the extremely high reactivity of singlet oxygen and its short lifetime in biological microenvironments.1–3 This can limit the chances of luminescence emission. Furthermore, current NIR detectors have very low quantum efficiency (the ratio of incident photons to converted electrons).
During the past decade, there have been two main approaches for imaging singlet oxygen luminescence and attempting to correlate it with the biological response during PDT.4 The first method uses an NIR photomultiplier tube or an indium gallium arsenide (InGaAs) photodiode linear array combined with a scanning laser beam and sample.5–7 This approach may be subject to motion artifacts during the long scan time (of the order of minutes, even for low spatial resolution), which pose a serious concern in vivo. Furthermore, the treatment itself may cause changes to the singlet oxygen kinetics that are then missed or incorrectly recorded. The second approach uses an NIR camera with a set of bandpass filters to discriminate the singlet oxygen luminescence from background fluorescence and phosphorescence.8 However, the long acquisition time (40–50s per image) and the low signal-to-noise ratio likely preclude this method for PDT studies.
We have developed a novel configuration of the Xenics Model XEVA-1.7-320 NIR-sensitive InGaAs camera to image the photodynamically generated singlet oxygen luminescence, using custom-designed NIR optics to collect it (see Figure 1). We coupled a semiconductor laser into a quartz fiber optic diffuser for illumination and to optimize the collection optics, providing accelerated image acquisition with acceptable signal-to-noise ratio. To demonstrate the system performance, we recorded luminescence images with each of three custom-made bandpass filters (25nm full width at half-maximum) followed by a longpass filter that reduces the background noise. We subtracted the average of the 1215 and 1315nm images, pixel by pixel, from that recorded at 1270nm to obtain the singlet oxygen luminescence image.
Figure 1. Schematic diagram of the singlet oxygen luminescence imaging system. DSWC: Dorsal skinfold window chamber. LP: Longpass. BP: Bandpass. NIR: Near-IR.
To show that the system works in vivo, we used mice implanted with dorsal skinfold window chambers. With these, we imaged the singlet oxygen luminescence in blood vessels immediately after intravenous injection with the model photosensitizer Rose Bengal. The window chamber enables direct imaging of the blood vessels in a single layer of skin without intervening tissue that causes blurring. We delivered PDT light to the other side of the skin layer through the epidermis (see Figure 1). Figure 2(A) shows representative luminescence images of the blood vessels. For the control condition (no photosensitizer), the vessels appear dark because of the strong absorption of irradiation light by oxyhemoglobin, as compared to surrounding tissue. Immediately after the photosensitizer injection, the intensities of the luminescence images increased for all three bandpass filters, but especially for that of 1270nm. After background subtraction, we could clearly see the image of singlet oxygen luminescence in the blood vessels, showing that it was well above the tissue background level: see Figure 2(B).
Figure 2. (A) Luminescence images of mouse blood vessels in the DSWC model taken before injection with Rose Bengal photosensitizer (control), immediately after injection (photosensitization), and two minutes after the mice were euthanized (hypoxic conditions). The acquisition time for each image was 2s. The scale bar (500μm) can be used to estimate the diameter of the vessels. (B) Pixel intensity profile along the dashed lines for the wavelength band and for the calculated singlet oxygen luminescence signal for the photosensitization condition. a.u.: Arbitrary units.
We euthanized the mice by anesthetic overdose and took images two minutes afterwards, where we saw the oxygen dependence of the luminescence. The NIR signal significantly decreased at all wavelengths, but the 1270nm signal was not above the background, further validating the interpretation that the 1270nm peak is due to singlet oxygen luminescence: see Figure 2(A).
In summary, the novel configuration of an NIR-sensitive InGaAs camera and optimized light collection enables direct imaging of the singlet oxygen luminescence generated in blood vessels during PDT. With a 2s image integration time, the system is practical for many in vivo studies. Future work will investigate the dependence of the luminescence images under varying PDT treatment conditions, such as photosensitizer type and concentration, oxygen concentration, and light dose, to better understand the mechanisms of singlet oxygen generation during PDT. In parallel, we will also investigate the quantitative correlation between the intensities of singlet oxygen luminescence images of blood vessels and the corresponding vasoconstriction (narrowing of blood vessels) after PDT, which holds the potential for establishing singlet oxygen luminescence-based dosimetry in clinical treatments targeting the vasculature.
This work was supported by the National Natural Science Foundation of China (61275216, 61036014), the Fujian Provincial Natural Science Foundation (2011J06022), and the Program for Changjiang Scholars and Innovative Research Team in Universities (IRT1115).
Lisheng Lin, Huiyun Lin, Shusen Xie, Buhong Li
Fujian Normal University
Lisheng Lin is a PhD candidate in optical engineering.
Huiyun Lin is a lecturer working on PDT dosimetry. She received her PhD in optical engineering in 2012.
Shusen Xie is a professor and the director of the Institute of Laser and Optoelectronics Technology. His research focuses on fluorescence spectroscopy and imaging for early cancer diagnosis and tissue optics of human meridians. He has published more than 200 peer-reviewed journal articles on biomedical photonics.
Buhong Li is a professor working on optical monitoring for PDT dosimetry, in particular for the detection of singlet oxygen luminescence. He received his PhD in optical engineering from Zhejiang University in 2003, and has published more than 50 peer-reviewed journal papers.
Defu Chen, Ying Gu
Chinese People's Liberation Army General Hospital and Medical College
Defu Chen is studying for a PhD in electronic science and technology at the Beijing Institute of Technology.
Ying Gu holds an MD in clinical medicine from Tianjin Medical University and a PhD in laser medicine from Chinese PLA Medical College. Since 1993 she has been chair of the Laser Medical Department and professor of laser medicine. She has published more than 200 scientific papers in basic, applied, and clinical laser medicine, and is the inventor of vascular target PDT for port wine stain treatment, which has been successfully used in China for more than 20 years.
Brian C. Wilson
University of Toronto
Princess Margaret Cancer Centre
Brian Wilson is a professor and senior research scientist and has a long-term interest in PDT dosimetry. He pioneered the use of NIR luminescence spectroscopy for singlet oxygen measurements, and now directs a multidisciplinary translational program in biophotonics. He is the 2014 recipient of the SPIE Britton Chance Biomedical Award.
1. M. T. Jarvi, M. J. Niedre, M. S. Patterson, B. C. Wilson, Singlet oxygen luminescence dosimetry (SOLD) for photodynamic therapy: current status, challenges, and future prospects, Photochem. Photobiol. 82, p. 1198-1210, 2006.
2. S. Hackbarth, J. Schlothauer, A. Preuss, B. Röder, New insights to primary photodynamic effects—singlet oxygen kinetics in living cells, J. Photochem. Photobiol. B. 98, p. 173-179, 2010.
3. B. Li, H. Lin, D. Chen, B. C. Wilson, Y. Gu, Singlet oxygen detection during photosensitization, J. Innov. Opt. Health Sci. 6, p. 1330002, 2013.
4. L. Lin, H. Lin, D. Chen, L. Chen, M. Wang, S. Xie, Y. Gu, B. C. Wilson, B. Li, Direct imaging of singlet oxygen luminescence generated in blood vessels during photodynamic therapy, Proc. SPIE
9129, p. 912920, 2014. doi:10.1117/12.2052033
5. L. M. Andersen, Z. Gao, P. R. Ogilby, L. Poulsen, I. Zebger, A singlet oxygen image with 2.5μm resolution, J. Phys. Chem. A. 106, p. 8488-8490, 2002.
6. M. J. Niedre, M. S. Patterson, A. Giles, B. C. Wilson, Imaging of photodynamically generated singlet oxygen luminescence in vivo, Photochem. Photobiol. 81, p. 941-943, 2005.
7. J. C. Schlothauer, J. Falckenhayn, T. Perna, S. Hackbarth, B. Röder, Luminescence investigation of photosensitizer distribution in skin: correlation of singlet oxygen kinetics with the microarchitecture of the epidermis, J. Biomed. Opt.
18, p. 115001, 2013. doi:10.1117/1.JBO.18.11.115001
8. S. Lee, M. E. Isabelle, K. L. Gabally-Kinney, B. W. Pogue, S. J. Davis, Dual-channel imaging system for singlet oxygen and photosensitizer for PDT, Biomed. Opt. Express 2, p. 1233-1242, 2011.