BOLD MRI: a tool for predicting tumor therapy outcome based on tumor blood oxygenation and vascular function

Ralph P Mason^*

The University of Texas Southwestern Medical Center, 5323 Harry Hines Boulevard, Dallas, TX 75390‑9058, USA

*Corresponding Author:

Ralph P Mason
The University of Texas Southwestern Medical Center
5323 Harry Hines Boulevard, Dallas
TX 75390‑9058, USA
Tel: +1 214 648 8926
Fax:+1 214 648 2991
E-mail:ralph.mason@ utsouthwestern.edu

Abstract

We believe that oxygen-sensitive MRI can be used to evaluate tumor hypoxia so that therapy may be individualized and optimized for the characteristics of an individual patient. Specifically, we foresee a simple test based on breathing hyperoxic gas to characterize tumor hypoxia and the potential for modulating tumor hypoxia and hence therapeutic response.

Radiology has become the mainstay for clinical diagnosis and together with pathology underpins clinical medicine. Imaging has come a long way from the revolutionary, but crude x‑rays of Roentgen at the turn of the 20th century to the exquisite temporal and spatial revolution offered by modern computed tomography, providing multislice (or 3D) sub-millimeter spatial and sub-second temporal resolution. Beyond x‑rays, MRI has revolutionized soft-tissue analysis by providing tissue contrast based on multiple parameters identified by varying repetition and echo times and gradients to reveal T₁‑, T2‑, T₂^*‑, diffusion- and flow-weighted images. Nonetheless, the discipline hitherto remained essentially an anatomical analysis providing tumor location, size and spread.

A new vision foresees prognostic radiology, whereby tumors are not only identified with high sensitivity and specificity, but there is stratification of tumors according to expected response to therapy, as well as early evaluation of therapeutic efficacy. Clinically, we are privileged to have a wide range of therapeutic choices, but this presents the challenge of selecting the right therapy. Societally, we are demanding greater efficacy at reduced cost. Increasingly, it has been realized that tumors that may outwardly appear quite similar may exhibit very different responses to therapy. As a result, there is increasing emphasis on tumor characterization using genomic, proteomic and metabolic techniques with the promise of assessing tumor aggressiveness and predicting response to therapy: generally, they require biopsy. A notable example is the evaluation of Her2‑neu expression for efficacy of Herceptin®. Noninvasive assessments are even more attractive.

One tumor characteristic that has come to the forefront of attention is tumor hypoxia and, more generally, oxygenation. Since the fundamental studies of Gray et al. [1] 50 years ago, it has been accepted that cellular response to radiation is oxygen dependant with a threefold increase in radiation resistance when pO₂ declines from approximately 10 mmHg to hypoxia (<1 mmHg). The original work used cultured cells in vitro, but noting the inefficient tortuous nature of tumor vasculature hypoxia was expected in vivo. A meta-analysis of some 10,000 patients indicated a marginal clinical benefit for manipulating tumor hypoxia [2]. However, the overall conclusion was that there was a pressing need to identify those patients (viz. tumors) who would actually benefit from the added cost and complexity of addressing hypoxia. The importance and relevance of hypoxia and feasibility of imaging hypoxia in tumors are drawing increasing interest. There is a consensus that while many methods have been proposed there is currently no gold standard available for assessing tumor hypoxia in patients. Criteria considered important for a useful technique include invasiveness, radiation exposure, resolution, safety, availability, imaging time and lead time to effective implementation in the clinic [3].

The development of the Eppendorf electrode system permitted measurement of pO₂ (mean, median, hypoxic percentiles and distributions) in vivo and several studies have now shown distinctly poorer prognosis for patients with large hypoxic tumors in the cervix and head and neck [4,5]. Moreover, hypoxia has been reported in the brain, lung and prostate. However, the needle electrode was sufficiently invasive to discourage widespread use and it is no longer manufactured. Fluoromisonidazole and similar agents such as EF5 based on 18F as well as 64Cu‑ATSM have been applied in clinical trials and promise insight into identifying tumor hypoxia using PET [6]. However, the associated radioactivity and inability to undertake rapid dynamic studies are limiting factors. 19F MRI efficiently measures pO₂ in preclinical studies based on perfluorocarbon-reporter molecules such as hexafluorobenzene and fluorocarbon relaxometry using echo planar imaging for dynamic oxygen mapping (FREDOM) has been successfully used to assess tumor hypoxia and interventions to modulate radiotherapy in rats [7]. Based on FREDOM, we have found three categories of tumor: category I hypoxic tumors, which do not respond to hyperoxic gas intervention; category II hypoxic tumors, which do respond to hyperoxic gas intervention; and category III, well-oxygenated tumors. In every case examined to date, well-oxygenated tumors became even better oxygenated with hyperoxic gas breathing. However, the lack of ¹⁹F MRI in human scanners has stymied development.

“A new vision foresees prognostic radiology, whereby tumors are not only identified with high sensitivity and specificity, but there is stratification of tumors according to expected response to therapy, as well as early evaluation of therapeutic efficacy.”

Ideally, tumor oxygenation would be characterized noninvasively generating prognostic criteria for therapy without the need for exogenous reporter molecules. The fundamental reports of Thulborn [8] and Ogawa [9] suggested that blood could reveal vascular oxygenation based on T₂^*‑weighted MRI. Deoxyhemoglobin is paramagnetic and induces signal loss in T₂^*‑weighted images and blood-oxygen-level-dependant (BOLD) contrast proton nuclear magnetic resonance forms the basis of so-called functional MRI in brain activation studies, where it is thought to reflect changes in perfusion. In tumors, a BOLD response is more difficult to interpret, since it may also be influenced by blood flow, as investigated extensively by Howe et al. [10], who coined the term flow-and-oxygen-level-dependant (FLOOD) contrast. In addition, variation in vascular volume can introduce signal perturbation. In tumors, BOLD effects are often evaluated in response to a challenge such as hyperoxic gas breathing, and several studies have indicated a correlation with relative pO₂ [11–13]. There was always a positive correlation, but a 10% change in relative signal intensity in the BOLD experiment could correspond to an increase in pO₂ of less than 25 torr or more than 100 torr. We contend that either change would be radiobiologically pertinent. Once pO₂ exceeds 10 torr there is relatively little further oxygen enhancement achieved by increasing pO₂ further. It has been suggested that BOLD measurements are only useful in wellvascularized tumors since the changes depend on deoxyhemoglobin concentration. However, we believe that the ability to oxygenate a tumor depends strongly on vascular supply and that a small BOLD response due to low vascularization will itself be indicative of hypoxic tumors.

Nonetheless, BOLD changes alone may not be definitive for tumor hypoxia. Recently, Matsumoto et al. showed that a T₁ response accompanied breathing hyperbaric oxygen for squamous cell carcinoma xenografts in mice [14]. The so-called tissue-oxygen-level-dependant (TOLD) response is sensitive to changes in tissue pO₂ , as opposed to vascular oxygenation. Since many additional factors such as protein concentrations and ions may influence the absolute T₁ value, it is not a reliable quantitative assay of absolute pO₂ , but we are finding it to be a robust surrogate of changes in pO₂ in response to interventions.

We propose that combined assessment of T₁ and T₂^* changes in tumors accompanying breathing hyperoxic gas will allow noninvasive characterization of tumor hypoxia. In preliminary investigations we have identified distinct characteristics for tumor oxygenation and dynamics: category I tumors show essentially no signal response to breathing hyperoxic gas; category II tumors show modest response; and category III show large BOLD and TOLD responses. As expected, BOLD responses precede TOLD since vascular oxygen delivery is required before changes in tissue pO₂.

We are already able to examine patients and have preliminary data in breast, cervix, head and neck and lung tumors. Lung tumors are most challenging owing to respiratory and cardiac motion. We should note that others are also pursuing such measurements, notably Padhani et al. (Mount Vernon, London, UK) in prostate based on T₂^* [15] and O’Connor et al. (Manchester, UK) characterizing TOLD in multiple organs and tissues [16]. We note that Rodrigues et al. reported relationships between baseline T₂^* and response to radiation [17].

Whichever oxygen-sensitive MRI method is ultimately found to be most useful, it has important implications for the practice of radiation oncology. Current radiotherapy uses extended fractionation regimens requiring patients to attend daily for many weeks. This is predicated on the need to avoid normal tissue damage. The development of highly focused beams and stereotactic body radiation therapy allows higher doses with a few fractions to be considered again. Hypoxia is likely to play a greater role for small numbers of high-dose irradiations, since there is less opportunity for the interfraction reoxygenation encountered between daily low doses. Each method of evaluating hypoxia presents new opportunities for individualized treatment exploiting ‘dose painting’ or intensity-modulated radiotherapy or a hypoxia-selective cell cytotoxin.

In terms of hypoxia in human tumors, we perceive twin needs: to develop an effective method of detecting hypoxia noninvasively prior to therapy, and to assess the ability to modulate tumor hypoxia in order to overcome resistance to therapy. We believe we are at an historic juncture where we not only have technologies for identifying hypoxia, but, more importantly, methods of tailoring therapy successfully to accommodate or exploit the hypoxia. In conclusion, BOLD can provide insight into tumor hypoxia, but we believe dynamic oxygen challenge evaluated by nuclear magnetic resonance T₁ and T₂^* (DOCENT), which exploits BOLD and TOLD contrast to noninvasively detect changes in tumor oxygenation using proton MRI, will provide a robust prognostic test to reveal tumor hypoxia.

Financial & competing interests disclosure

Observations and opinions discussed in this article are based on investigations supported in part by grants from the Susan G Komen Foundation, The Mary Kay Ash Foundation and NIH U24CA126608. The author has no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.

No writing assistance was utilized in the production of this manuscript.

References

1. Gray L, Conger A, Ebert M et al.: The concentration of oxygen dissolved in tissues at time of irradiation as a factor in radiotherapy. Br. J. Radiol. 26, 638-648 (1953)
2. Overgaard J, Horsman MR: Modification of hypoxia-induced radioresistance in tumors by the use of oxygen and sensitizers. Semin. Radiat. Oncol. 6, 10–21 (1996).
3. Tatum JL, Kelloff GJ, Gillies RJ et al.: Hypoxia: importance in tumor biology, noninvasive measurement by imaging, and value of its measurement in the management of cancer therapy. Int. J. Radiat. Biol. 82, 699– 757 (2006).
4. Fyles A, Milosevic M, Pintilie M et al.: Long-term performance of interstial fluid pressure and hypoxia as prognostic factors in cervix cancer. Radiother. Oncol. 80, 132–137 (2006).
5. Brizel DM, Sibly GS, Prosmitz LR et al.: Tumor hypoxia adversely affects the prognosis of carcinoma of the head and neck. Int. J. Radiat. Oncol. Biol. Phys. 38, 285–289 (1997).
6. Krohn KA, Link JM, Mason RP: Molecular imaging of hypoxia. J. Nucl. Med. 49(Suppl. 2), 129S–148S (2008).
7. Zhao D, Jiang L, Mason RP: Measuring changes in tumor oxygenation. Meth. Enzymol. 386, 378–418 (2004).
8. Thulborn KR, Waterton JC, Matthews PM et al.: Oxygenation dependence of the transverse relaxation time of water protons in whole blood at high field. Biochim. Biophys. Acta 714, 265–270 (1982).
9. Ogawa S, Lee TM, Kay AR et al.: Brain magnetic resonance imaging with contrast dependent on blood oxygenation. Proc. Natl Acad. Sci. USA 87, 9868–9872 (1990).
10. Howe FA, Robinson SP, Rodrigues LM et al.: Flow and oxygenation dependent (FLOOD) contrast MR imaging to monitor the response of rat tumors to carbogen breathing. Magn. Reson. Imaging 17, 1307–1318 (1999).
11. Baudelet C, Gallez B: How does blood oxygen level-dependent (BOLD) contrast correlate with oxygen partial pressure (pO₂) inside tumors? Magn. Reson. Med. 48, 980–986 (2002).
12. Fan X, River JN, Zamora M et al.: Effect of carbogen on tumor oxygenation: combined fluorine-19 and proton MRI measurements. Int. J. Radiat. Oncol. Biol. Phys. 54, 1202–1209 (2002).
13. Zhao D, Jiang L, Hahn EW, Mason RP: Comparison of 1H BOLD and 19F MRI to investigate tumor oxygenation. Magn. Reson. Med. 62, 357–364 (2009).
14. Matsumoto K, Bernardo M, Subramanian S et al.: MR assessment of changes of tumor in response to hyperbaric oxygen treatment. Magn. Reson. Med. 56, 240–246 (2006).
15. Alonzi R, Padhani AR, Maxwell RJ et al.: Carbogen breathing increases prostate cancer oxygenation: a translational MRI study in murine xenografts and humans. Br. J. Cancer 100, 644–648 (2009).
16. O’Connor JPB, Jackson A, Buonaccorsi GA et al.: Organ-specific effects of oxygen and carbogen gas inhalation on tissue longitudinal relaxation times. Magn. Reson. Med. 58, 490–496 (2007).
17. Rodrigues LM, Howe FA, Griffiths JR et al.: Tumor R2* is a prognostic indicator of acute radiotherapeutic response in rodent tumors. J. Magn. Reson. Imaging 19, 482-488 (2004).