Elsevier

Methods in Enzymology

Volume 386, 2004, Pages 349-378
Methods in Enzymology

Near-Infrared Spectroscopy and Imaging of Tumor Vascular Oxygenation

https://doi.org/10.1016/S0076-6879(04)86017-8Get rights and content

Publisher Summary

This chapter elaborates the near-infrared spectroscopy and imaging of tumor vascular oxygenation. A large number of investigations have been conducted in both laboratory and clinical settings to noninvasively monitor tissue vascular oxygenation using near-infrared (NIR) spectroscopy and imaging. The NIR studies of the brain include detection of brain injury or trauma, determination of cerebrovascular hemodynamics and oxygenation, and functional brain imaging in response to a variety of neurologic activations. The elevated oxygenation process is completely reversible upon returning air breathing, but still present 10 to 20 min after the baseline inhalation in many cases. The high reproducibility of results suggests that one can apply repeated interventions to explore the efficacy of interventions designed to alter tumor vascular oxygenation. It is found that a combination of well-perfused and poorly perfused mechanisms in tumor vasculature will result in the coexistence of two time constants. It is found that the concentration changes in oxygenated hemoglobin measured from both breast and prostate tumors often display a very prompt rise, followed by a gradual persistence throughout the intervention.

Introduction

In recent years, a large number of investigations have been conducted in both laboratory and clinical settings to noninvasively monitor tissue vascular oxygenation using near-infrared (NIR) spectroscopy (NIRS) and imaging. Although the NIR imaging techniques are limited by their spatial resolutions, they have a great potential to be developed as a new imaging modality because of their capabilities to provide functional images. NIR spectroscopy and imaging research has been mainly focused on two organs: (1) the brain and (2) the breast.

The NIR studies of the brain include detection of brain injury or trauma,1 determination of cerebrovascular hemodynamics and oxygenation,2, 3 and functional brain imaging in response to a variety of neurologic activations.4, 5, 6, 7 NIR functional brain imaging increasingly has become of great interest in studying hemodynamic response to brain activation.8 This is mainly because the optical signals of the NIR techniques are able to noninvasively penetrate through the scalp and skull of an adult human and are sensitive to changes in the concentration of oxygenated (HbO) and deoxygenated hemoglobin (Hb). Although it is difficult to obtain very accurate quantification of cerebral HbO and Hb concentrations from the NIR imaging techniques due to rigorous requirements of theory and boundary conditions,9 the techniques can offer relatively accurate measurements of changes of HbO and Hb, thus providing quantitative changes in total cerebral blood volume (CBV, which is assumed to be proportional to total hemoglobin concentration, HbT, where HbT = HbO + Hb).

The objective of NIR breast imaging is to develop a novel functional imaging modality for early breast cancer detection and diagnosis beyond currently available techniques. Various efforts by several research groups10, 11, 12, 13, 14, 15 have been made in either laboratory or clinical studies. For example, the research groups of Paulsen and colleagues12 and of Jiang15 have developed a frequency-domain (FD) 16-source, 16-detector breast imager and have reported their in vivo results of optical properties of abnormalities from female volunteers and patients. The research group led by Chance has employed a time-domain (TD) 32-channel imaging system in conjunction with magnetic resonance imaging (MRI) to increase specificity and sensitivity of breast cancer detection.13 Because of simplicity and low cost in comparison to the FD and TD imaging systems, continuous-wave (CW) NIR breast imaging systems also have been developed by the groups of Barbour et al.14 and Chance.16 These systems are currently under clinical tests for better breast cancer detection and diagnosis.

Although Hb, HbO, and HbT concentrations and light-scattering properties of tumors may be different from those of surrounding tissues, the optical contrast between tumor and surrounding tissue is about 2–3-fold at most in absorption, and much less in light scattering.17 Thus much effort on increasing the optical contrast between tumor and healthy surrounding tissues also has been made using fluorescence imaging18 or molecular beacons16 to detect and diagnose cancer or tumor with improved sensitivity and specificity.

However, efforts in using NIR techniques to monitor tumor responses to therapeutic interventions19, 20, 21 and therapy, such as to chemotherapy22 or photodynamic therapy,23 are very limited and preliminary. Moreover, few reports have been found on using NIR spectroscopy and imaging as a prognostic tool for therapy planning and optimization, or for tumor prognosis. Based on the existing knowledge and development of NIR spectroscopy and imaging for the brain and breast, we were motivated a few years ago by the possibility of using NIRS as an efficient, real-time, noninvasive means to monitor tumor vascular oxygenation during respiratory interventions. The recent results in our study were obtained from animal breast and prostate tumors in vivo with a one-channel NIR spectrometer. The data have demonstrated that the NIR techniques could have applications as a prognostic means accompanying cancer therapy.19, 20, 24

The presence and significance of tumor hypoxia has been recognized since the 1950s. Hypoxic cells in vitro and in animal tumors in vivo are documented to be three times more resistant to radiation-induced killing compared with aerobic cells.25 Recent studies have shown that tumor hypoxia is a possible prognostic indicator and is related to the aggressiveness of a tumor, the clinical stage, and poor clinical outcome.26, 27, 28, 29, 30 To improve the efficacy of oxygen-dependent treatment modalities, one possible strategy is the reduction of tumor hypoxia by raising the arterial oxygen partial pressure, PaO2 to overcome diffusion-limited hypoxia. Raising the arterial pO2 by breathing hyperoxic or hyperbaric gas mixtures could be effective.31, 32 However, attempts to apply increased oxygen breathing in the clinic have not always been successful, and this may be attributed to the inability to identify those patients with hypoxic tumors.33 More specifically, it may be attributed to the inability to identify those patients who would benefit from such interventions. Therefore evaluation of tumor oxygenation distributions and their changes during various stages of tumor growth, during therapeutic interventions, and during therapy is needed. Such evaluation can provide a better understanding of tumor development and tumor response to therapy, potentially allowing therapy to be planned to individual characteristics.

Numerous studies on tumor oxygen tension (pO2) measurements have been conducted in recent years using a variety of methods,34, 35, 36 such as microelectrodes,37 optical reflectance,38 electron spin resonance (EPR),39, 40 or MRI.36, 41, 42, 43, 44, 45 The latter two methods have advantages of making repeated measurements of pO2 noninvasively, and MRI has the further advantage of providing dynamic maps of pO2, which can reveal tumor heterogeneity.46 Although NIRS does not quantify pO2, it can indicate dynamic changes in vascular oxygenation and has the advantage of being entirely noninvasive, providing real-time measurements, and being cost-effective and portable. Our recent studies have revealed the need for NIR imaging of tumor vasculatures to study tumor heterogeneous response to therapeutic interventions and therapy. In the following sections, we briefly introduce basic algorithms used to quantify tumor hemoglobin oxygenation, followed by the NIR instrument description, system calibration, and experimental methods. Then, we provide several representative results taken from both prostate and breast tumors under hyperoxic respiratory interventions, using both one-channel and multichannel NIR systems, as well as one-channel and three-channel pO2 fiber optic needle probes. At the end, we wish to demonstrate that the NIR techniques are complementary with tumor pO2 readings and can be used as a new prognostic means for cancer therapy prognosis and therapy planning.

Section snippets

Theory and Algorithms

It is well known that hemoglobin concentrations and oxygen saturation in tissue vasculature can be determined using NIRS, since light absorptions of HbO and Hb in the NIR range are distinct. As with our previous work,19, 20 we assumed that HbO and Hb are the major significant absorbing species in tissue vasculature, including tumors, within the selected NIR range of 700–900 nm. Although diffusion theory has been a well-accepted theoretical approach to mathematically quantify light–tissue

One-Channel and Multichannel NIR System

A dual-wavelength (at 758 and 785 nm), one-channel NIR system (NIM, Inc., Philadelphia, PA) uses an in-phase and quadrature-phase chip (IQ chip). As shown in Fig. 1, the NIR system starts with a radiofrequency (RF) source to modulate the light intensities of two laser diodes (LD1 and LD2) at 140 MHz through a time-sharing system. The light passes through a bifurcated fiber optic probe, is transmitted through the tumor tissue, and is collected by a second fiber bundle. The light is then

HbO and pO2 Changes in Prostate Tumors Under Carbogen Intervention

We have measured relative changes of [HbO], [Hb]total, and tumor tissue pO2 from several Dunning prostate R3327-HI tumors, and Fig. 7 shows two representative data sets. Figure 7A shows the temporal profiles of Δ[HbO] and pO2 in a Dunning prostate R3327-HI tumor (3.6 cm3) measured simultaneously with the NIRS and pO2 needle electrode during carbogen respiratory challenge. After the breathing gas was switched from air to carbogen, Δ[HbO] increased rapidly, whereas Δ[Hb]total seemed to have much

Discussions

NIRS is noninvasive and provides a real-time assessment of changes in tumor vascular hemoglobin oxygenation. In this chapter, we basically provide demonstration of the ease and utility of NIRS studies of tumors. By switching the inhaled gas from air to carbogen, the NIRS measurement produces a rapid biphasic elevation in Δ[HbO]. The rapid time constant is in the range of seconds to a minute, whereas the slow component (10 to 50 times slower) continues for many minutes (see Fig. 7, Fig. 8). The

Conclusions

In conclusion, we believe that NIRS presents a new potential imaging modality to examine tumor vasculature rapidly, noninvasively, and cost-effectively. Ease of implementation and operation permit rapid application to accessible tumors in cancer patients. The inherent compatibility of fiber-optics technology and light with other modalities, such as electrodes20, 24 and MRI,71 will facilitate multiparametric multimodality investigations of tumor heterogeneity and vasculature in the near future.

Acknowledgements

This work was supported in part by the Department of Defense Breast Cancer Research grants BC990287 (HL) and BC000833 (YG), and NIH R01 CA79515 (NCI)⧸EB002762 (NIBIB) (RPM). We are grateful to Vincent Bourke for his collaborative work on multichannel pO2 measurements and Dr. Anca Constantinescu for her assistance with all the tumor investigations. We also gratefully acknowledge Dr. Britton Chance for his technical support on the multichannel NIR system.

References (71)

  • D.A. Boas et al.

    Neuroimage

    (2001)
  • B. Tromberg et al.

    Neoplasia

    (2000)
  • A.W. Fyles et al.

    Radiother. Oncol.

    (1998)
  • O. Thews et al.

    Radiother. Oncol.

    (2002)
  • D. Zhao et al.

    Methods Enzymol.

    (2004)
  • D. Le et al.

    Magn. Reson. Imaging

    (1997)
  • S. Hunjan et al.

    Int. J. Radiat. Oncol. Biol. Phys.

    (1998)
  • D. Zhao et al.

    Neoplasia

    (2003)
  • M. Ferrari et al.

    J. Photochem. Photobiol.

    (1992)
  • E.W. Hahn et al.

    Magn. Reson. Imaging

    (1993)
  • R.P. Mason et al.

    Int. J. Radiat. Oncol. Biol. Phys.

    (1994)
  • X. Fan et al.

    Int. J. Radiat. Oncol. Biol. Phys.

    (2002)
  • M. Kragh et al.

    Eur. J. Cancer

    (2001)
  • B.M. Fenton

    Radiother. Oncol.

    (1997)
  • S. Gopinath et al.

    J. Neurosurg.

    (1993)
  • C. Cheung et al.

    Phys. Med. Biol.

    (2001)
  • G. Zhang et al.

    Phys. Med. Biol.

    (2000)
  • M. Fabiani et al.

    J. Biomedi. Opti.

    (1996)
  • R. Wenzel et al.

    J. Biomed. Opt.

    (1996)
  • B. Chance et al.

    Opt. Express.

    (1998)
  • D. A. Boas, G. Jasdzewski, G. Strangman, J. P. Culver, and R. Poldrack, “OSA Biomedical Topical Meetings, Technical...
  • J. C. Hebden, E. M. C. Hillman, A. Gibson, N. Everdell, R. M. D. Yusof, D. T. Delpy, S. R. Arridge, T. Austin, and J....
  • S. Fantini et al.

    Appl. Opt.

    (1998)
  • T.O. McBride et al.

    Opt. Lett.

    (2001)
  • V. Ntziachristos et al.
  • R. L. Barbour, C. H. Schmitz, H. L. Graber, and Y. Pei, “OSA Biomedical Topical Meetings, Technical Digest,” p. 456....
  • X. Gu, N. Iftimia, Y. Xu, H. Jiang, and L. L. Fajardo, “OSA Biomedical Topical Meetings, Technical Digest,” p. 468....
  • B. Chance, “OSA Biomedical Topical Meetings, Technical Digest” p. 450. Miami, FL, April 7–10,...
  • J.B. Fishkin et al.

    Appl. Opt.

    (1997)
  • R.H. Mayer et al.

    Appl. Opt.

    (1999)
  • H. Liu et al.

    Appl. Opt.

    (2000)
  • J.G. Kim et al.

    J. Biomed. Opt.

    (2003)
  • E.L. Hull et al.

    Br. J. Cancer

    (1999)
  • D. B. Jakubowski, A. E. Cerussi, F. Bevilacqua, N. Shah, and B. J. Tromberg, “OSA Biomedical Topical Meetings,...
  • H. Wang, T. C. Zhu, M. Solonenko, S. M. Hahn, J. Metz, A. Dimofte, J. Mile, A. G. Yodh, “OSA Biomedical Topical...
  • Cited by (0)

    View full text