Magnetic resonance in the era of molecular imaging of cancer

Abstract

Magnetic resonance imaging (MRI) has played an important role in the diagnosis and management of cancer since it was first developed, but other modalities also continue to advance and provide complementary information on the status of tumors. In the future, there will be a major continuing role for noninvasive imaging in order to obtain information on the location and extent of cancer, as well as assessments of tissue characteristics that can monitor and predict treatment response and guide patient management. Developments are currently being undertaken that aim to provide improved imaging methods for the detection and evaluation of tumors, for identifying important characteristics of tumors such as the expression levels of cell surface receptors that may dictate what types of therapy will be effective and for evaluating their response to treatments. Molecular imaging techniques based mainly on radionuclide imaging can depict numerous, specific, cellular and molecular markers of disease and have unique potential to address important clinical and research challenges. In this review, we consider what continuing and evolving roles will be played by MRI in this era of molecular imaging. We discuss some of the challenges for MRI of detecting imaging agents that report on molecular events, but highlight also the ability of MRI to assess other features such as cell density, blood flow and metabolism which are not specific hallmarks of cancer but which reflect molecular changes. We discuss the future role of MRI in cancer and describe the use of selected quantitative imaging techniques for characterizing tumors that can be translated to clinical applications, particularly in the context of evaluating novel treatments.

Figure 1

The growth of molecular imaging in the past decade as measured by the number of publications each year reported in the PubMed database.

Figure 2

(Left) Transaxial cross section of human brain acquired in 9 minutes at 0.15 Tesla in 1984 (voxel sizes ≈ 2 × 2 × 8 mm) (Right) part of the same mid-brain region imaged at 7 Tesla in 2008 in 4 minutes (voxel size 0.5 × 0.5 × 3 mm)

Figure 3

a selection of different types of MR images produced by a standard, modern clinical scanner. From top to bottom, left to right these are (a) T1 weighted (b) T2 weighted (c) a map of quantitative magnetization transfer pool ratio (d) blood perfusion (e) a BOLD map of activation (f) a venogram (g) diffusion weighted (h) a DTI image of white matter tracts (i) multi-voxel high resolution MR spectra

Figure 4

ADC maps overlaid on T2-weighted images of a mouse model of HER2+ breast cancer treated with trastuzumab; (Top). at baseline and (Bottom) one week (two cycles) after treatment. Residual tumor ADC values generally increase after successful treatment with the drug.

Figure 5

DCE imaging of human breast cancer. (Top row) K^trans maps overlaid on T₁-weighted SPGRE image acquired before treatment (panel a), after one cycle of neoadjuvant chemotherapy (panel b) and at the conclusion of a therapy but prior to surgery. This is an example of patient showing a complete clinical response. (Bottom row) Similar data for a patient exhibiting progressive disease. In the top panel, there is a general decrease in K^trans values from pre- to post-one cycle of therapy (panels a and b, respectively), whereas in the bottom row there is a general increase (panels d to e) and this correlates with disease burden at the time of therapy.