Imaging the Tumor Microenvironment
Geoff J M Parker1
1Centre for Medical Image Computing, University College London, London, United Kingdom

Synopsis

MRI is able to provide measurements of a range of tumour microenvironmental features. This lecture will summarise these capabilities and provide detailed examples, including measurements of tumour microstructure using diffusion MRI and measurements of hypoxia using an oxygen challenge. Example applications of these measurements will be presented and the importance of validation of MRI measurements will be discussed.

Target audience

This lecture will benefit researchers with an interest in some of the current methods for assessing the tumour microenvironment using MRI.

Outcomes

By attending this lecture, participants should be able to:
  • Appreciate the value of quantification of microenvironmental features in cancer
  • Appreciate the range of microenvironmental features that may be accessed using MRI
  • Understand the information available from diffusion MRI of tumours
  • Understand the information available from MRI-based oxygenation measurements in tumours
  • Understand the importance of validation of MRI biomarkers for assessing the tumour microenvironment

The role of MRI in assessing the tumour microenvironment

MRI has two major advantages over most imaging modalities: it is largely non-invasive, allowing repeat scanning, and it can be sensitised to a wide range of structural, physiological, and metabolic characteristics of tumours. These properties make it well-suited to probing the tumour microenvironment, with multiple measurement options available. For example: angiogenesis can be monitored using dynamic contrast-enhanced MRI to provide information on perfusion and capillary permeability (e.g. 1) and vessel size imaging to provide information on microvessel geometry (e.g. 2); pH can be probed using CEST and other methods (e.g. 3); hypoxia can be measured using susceptibility contrast (e.g. 4) or using T1 contrast (e.g. 5); and cell packing and geometry information can be derived from suitably-acquired diffusion MRI acquisitions (e.g. 6). Within reasonable time limits, multiple measurements can be deployed in an imaging session, providing potentially comprehensive information on the microenvironment. Deployment in longitudinal studies sheds light on how the microenvironment changes during tumour growth and in response to intervention (e.g. 7).

Example: Diffusion MRI for quantification of tumour microstructure

An example of a microenvironmental measurement that continues to receive widespread interest is the use of diffusion MRI to enable estimation of cell size and packing density. It is expected that cell density will be higher in tumours than in healthy tissue, which leads to the diagnostic utility of diffusion-weighted measurements and simple parameterisations, such as the estimation of the apparent diffusion coefficient (ADC). During cancer treatment, the defining event is the death of tumour cells, a process that can lead transiently to tumour cells reducing in size without an overall change in the number of cells, (for example, in cells undergoing apoptotic cell shrinkage). Subsequently a reduced density of cells may be present, with those remaining likely to be of similar size to those present before treatment. Importantly, both scenarios may cause an increase in ADC (e.g. 8), suggesting that more specific measurements of cell size and cell packing may be of relevance in understanding response to therapy.

Example: Detecting tumour hypoxia using oxygen-enhanced MRI

Tumours often display reduced levels of oxygen tension due to the combination of poor perfusion characteristics and high metabolic demand. Such hypoxic conditions can lead to increased angiogenesis, metastasis and radioresistance; quantification of hypoxia is therefore of interest for understanding tumour growth and likely response to treatment. The use of T1 contrast to monitor the impact of increased concentrations of inhaled oxygen (oxygen-enhanced MRI (e.g. 9), sometimes also known as tissue oxygenation level-dependent (TOLD) contrast (e.g. 10)) has been shown to quantify the extent of hypoxia in a number of tumour models. This approach has also been shown to be able to track changes in hypoxic profile in response to hypoxia-targeting drugs (11) and radiotherapy (12), and to predict the growth delay due to radiotherapy (10).

The importance of validation

Despite the wide range of capabilities provided by MRI, quantification can be challenging due to MRI's sensitivity to unwanted effects ranging from image artefacts to sensitivity to 'off-target' phenomena. In order to understand the specificity of a measurement to the characteristic of interest it is essential to perform comprehensive validation against a gold-standard approach. Preclinical MRI provides excellent opportunities for comparison of MRI measurements with histology, other imaging such as high resolution CT, nuclear medicine, or high frequency ultrasound, and other forms of invasive sampling, such as probes of oxygen tension or pH.

Acknowledgements

With thanks to James O'Connor and Damien McHugh for the provision of materials.

References

1. Zhang, J., Feng, L., Otazo, R., & Kim, S. G. (2019). Rapid dynamic contrast-enhanced MRI for small animals at 7T using 3D ultra-short echo time and golden-angle radial sparse parallel MRI. Magnetic Resonance in Medicine, 81(1), 140–152. https://doi.org/10.1002/mrm.27357

2. Burrell, J. S., Bradley, R. S., Walker-Samuel, S., Jamin, Y., Baker, L. C. J., Boult, J. K. R., … Robinson, S. P. (2012). MRI measurements of vessel calibre in tumour xenografts: Comparison with vascular corrosion casting. Microvascular Research, 84(3), 323–329. https://doi.org/10.1016/j.mvr.2012.08.001

3. Chen, L. Q., & Pagel, M. D. (2015). Evaluating pH in the Extracellular Tumor Microenvironment Using CEST MRI and Other Imaging Methods. Advances in Radiology, 2015, 1–25. https://doi.org/10.1155/2015/206405

4. Robinson, S. P., Kalber, T. L., Howe, F. A., McIntyre, D. J. O., Griffiths, J. R., Blakey, D. C., … Waterton, J. C. (2005). Acute Tumor Response to ZD6126 Assessed by Intrinsic Susceptibility Magnetic Resonance Imaging. Neoplasia, 7(5), 466–474. https://doi.org/10.1593/neo.04622

5. O’Connor, J. P. B., Boult, J. K. R., Jamin, Y., Babur, M., Finegan, K. G., Williams, K. J., … Robinson, S. P. (2016). Oxygen-Enhanced MRI Accurately Identifies, Quantifies, and Maps Tumor Hypoxia in Preclinical Cancer Models. Cancer Research, 76(4), 787–795. https://doi.org/10.1158/0008-5472.CAN-15-2062

6. Panagiotaki, E., Walker-Samuel, S., Siow, B., Johnson, S. P., Rajkumar, V., Pedley, R. B., … Alexander, D. C. (2014). Noninvasive Quantification of Solid Tumor Microstructure Using VERDICT MRI. Cancer Research, 74(7), 1902–1912. https://doi.org/10.1158/0008-5472.CAN-13-251

7. McHugh, D. J., Lipowska‐Bhalla, G., Babur, M., Watson, Y., Peset, I., Mistry, H. B., … Parker, G. J. M. (2020). Diffusion model comparison identifies distinct tumor sub‐regions and tracks treatment response. Magnetic Resonance in Medicine, (Early view). https://doi.org/10.1002/mrm.28196

8. McHugh, D. J., Hubbard Cristinacce, P. L., Naish, J. H., & Parker, G. J. M. (2019). Towards a ‘resolution limit’ for DW-MRI tumor microstructural models: A simulation study investigating the feasibility of distinguishing between microstructural changes. Magnetic Resonance in Medicine, 81(4), 2288–2301. https://doi.org/10.1002/mrm.27551

9. O’Connor, J. P. B., Naish, J. H., Parker, G. J. M., Waterton, J. C., Watson, Y., Jayson, G. C., … Jackson, A. (2009). Preliminary study of oxygen-enhanced longitudinal relaxation in MRI: a potential novel biomarker of oxygenation changes in solid tumors. International Journal of Radiation Oncology, Biology, Physics, 75(4), 1209–1215.

10. Hallac, R. R., Zhou, H., Pidikiti, R., Song, K., Stojadinovic, S., Zhao, D., … Mason, R. P. (2014). Correlations of noninvasive BOLD and TOLD MRI with pO2 and relevance to tumor radiation response. Magnetic Resonance in Medicine, 71(5), 1863–1873. https://doi.org/10.1002/mrm.24846

11. Little, R. A., Tessyman, V., Babur, M., Cheung, S., Gieling, R., Finegan, K. G., … O’Connor, J. P. (2017). In vivo OE-MRI quantification and mapping of response to hypoxia modifying drugs Banoxantrone and Atovaquone in Calu6 xenografts. ISMRM 25, 2919.

12. Salem, A., Little, R. A., Latif, A., Featherstone, A. K., Babur, M., Peset, I., … O’Connor, J. P. B. (2019). Oxygen-enhanced MRI Is Feasible, Repeatable, and Detects Radiotherapy-induced Change in Hypoxia in Xenograft Models and in Patients with Non–small Cell Lung Cancer. Clinical Cancer Research, 25(13), 3818–3829. https://doi.org/10.1158/1078-0432.CCR-18-3932

Proc. Intl. Soc. Mag. Reson. Med. 28 (2020)