Dynamic Contrast-Enhanced Magnetic Resonance Imaging Calibration between Sites and Comparability
Sandeep N. Gupta
Dynamic contrast-enhanced (DCE) magnetic resonance imaging (MRI) is a method used to investigate the kinetics of injected low molecular weight contrast media distribution in tissue. DCE can be used for initial diagnosis, monitoring disease, and treatment. The primary use of DCE is in tumors, but DCE is useful in other applications where vascular permeability, flow, or volume are suspected to change, such as inflammation. In fact, DCE has the potential to become a quantitative biomarker in predicting outcome and monitoring of therapy response or pharmacodynamics of a drug, particularly in tumors.
Motivation for Use of DCE in Tumors
Recently, several novel target-specific anticancer drugs have been embedded at first- and second-line systemic therapy protocols. These target-specific therapies predominantly result in a “tumor stabilization” effect with changes in the tumor physiology, which will not necessarily be reflected by classical measurements of tumor size, such as Response Evaluation Criteria in Solid Tumors (see Chapter 56). Even if these cytostatic drugs result in measureable tumor shrinkage, response classification by these conventional criteria will still require several weeks (routinely, the first follow-up is acquired 6 to 8 weeks after treatment initiation). However, in cases that lack the specific target structure, tumor therapy will be ineffective and a swift adaptation to treatment is needed. Because these molecular targeting drugs are expensive in their preclinical development and daily clinical use, robust noninvasive biomarkers are strongly needed for early assessment of treatment response for drug discovery, patient care, and economic reasons.
Although they have been available for many years, DCE MRI techniques have enjoyed increasing popularity because of the availability of antiangiogenic drugs. Response of a patient to these drugs could be shown on DCE as early as 2 days after application of a first dose. Such therapies were used for brain, breast, prostate, and liver tumors.
The key to successful DCE imaging is standardized protocols if imaging was done at different centers (or at different magnets at one site) or with different protocols (Fig. 35.1).
DCE MRI acquisition is a multistep process. For the purposes of this chapter, it is assumed that the acquisition takes places at a qualified site, using a qualified scanner, and that it is understood where the regions are located that can be assessed beneficially by DCE MRI. For qualification criteria and processes see also QIBA-profile v1 (qibawiki.rsna.org.easyaccess1.lib.cuhk.edu.hk). The acquisition process should be timed with respect to the expected biologic effect due to an intervention of interest (e.g., a therapeutic intervention). DCE MRI does not require any specific preparation of the patient prior to the scan. Local requirements apply with regard to patient consent, administration of contrast media, and patient management according to standard of care for magnetic resonance procedures.
The typical imaging procedure consists of the following steps:
Anatomic sequences T1 weighted, T2 weighted
Variable flip angle (VFA) T1-weighted imaging (T1 mapping) or any other qualified tissue T1 mapping
2D/3D gradient echo volumetric imaging (dynamic imaging)
Anatomic, postcontrast T1-weighted imaging
TABLE 35.1 DETAILED SCAN PARAMETERS FOR 1.5T
A manufacturer-specific overview of detailed scan parameters for 1.5T can be found in Table 35.1. The use of T1 phantoms might be beneficial for T1 corrections but are not required as standard of care.
For tumors, the best suitable target lesion should be identified on an image used for actual tumor staging and should be based on tumor size (ideally diameter ≥2 cm), a location without pulsatile or respiratory artifacts, and solid tumor tissue with absence of extensive necrotic or cystic areas. Patient placement, coil positioning, and field of view should follow standard of care for the body region of interest (e.g., supine position for liver by using a torso coil and prone position for breast by using a breast coil). The use of parallel imaging should be handled with care due to potential effects from residual aliasing and regional noise variation (i.e., g-factor). However, parallel imaging is critical to improve temporal resolution. Particularly, with 3D acquisition, parallel imaging should be considered—in moderation—along both phase-encode dimensions.
When setting the protocol for DCE studies, the acquisition plane should include the lesion of interest and an arterial vessel in order to depict the arterial input function (AIF), as described in detail later. Acquisition planes that are used commonly for DCE MRI are for brain at transversal, for breast at sagittal, for prostate at transversal, and for liver at coronal orientation. For DCE MRI, as many slices as possible should be acquired while maintaining optimal temporal resolution. That is, for prostate 10 or more slices to cover the whole organ, whereas for liver 1 to 3 slices to enable high temporal resolution (without parallel imaging technique). In general, the temporal resolution should be less than 10 seconds, but ideally 5 seconds or less, especially for tumor locations prone to motion artifacts (e.g., liver). To reduce motion artifacts, patient positioning should be as comfortable as possible, relaxed breathing should be advised, and contrast media should be delivered by means of power injector, preferably at body temperature. However, for organs with strong motion artifacts, such as liver, postprocessing motion correction might be necessary. Image data reconstruction will be performed as specified by the respective vendor. All the acquisition process steps that have been mentioned are subject to quality control steps as outlined in (QIBA-profile v1).