Quantification Accuracy of Dynamic Contrast-Enhanced MRI for Prostate Cancer
Dynamic contrast-enhanced (DCE) MRI, when coupled with pharmacokinetic (PK) modelling of the contrast agent uptake curves, has shown considerable potential as a quantitative marker of tumour malignancy and also for monitoring of therapy. It involves the acquisition of a series of closely-spaced (in time) T1-weighted MR images as a bolus of gadolinium-based contrast agent (CA) is administered to a patient, and monitors the uptake of the CA into tissue (and tumour) and the subsequent wash-out from the tissue back into the blood stream. This results in a time-dependent signal time-intensity curve (STC) at each voxel within the imaged field of view, from which contrast time-intensity curves (CTCs) can be derived. The shape of these CTCs are reflective of the degree of CA uptake, which is related to the degree of malignancy based on the idea that tumours are fed by a dense network of leaky vasculature, from which the CA 'leaks' quickly into the surrounding tumour tissue (leading to a characteristic fast signal rise at early time points) followed by a 'wash-out' phase where the CA diffuses back into the blood stream. Healthy tissue, by comparison, demonstrates a slower initial signal increase since the CA is retained within the vasculature for longer, and later by a more gradual wash-out phase.
Semi-quantitative phenomenological parameters extracted from the STCs have shown some efficacy for differentiating malignant from healthy prostate tissue; such parameters include onset to enhancement, wash-in gradient, maximum intensity, wash-out gradient, and integral area under gadolinium contrast concentration after 60 seconds. However, pharmacokinetic modelling of CTC data offers the prospect of extracting and quantifying sub-voxel functional physiological information, such as the volume-transfer constant (Ktrans) and the fractional volume of the extravascular extracellular space (ve), allowing for a more accurate elucidation of tumour biology.
DCE-MRI has met with limited clinical uptake, in part due to the lack of sensitivity and specificity in separating tumour from healthy tissue. This can be seen in the published values of Ktrans and ve for many cancer types (whether prostate, breast, oesophageal, etc) which vary widely across research groups and ultimately cast doubt on the utility of the approach. While the precise reasons for this disparity are not clear, it is clear that the broad range of acquisition protocols which are commonly used to acquire the data is contributing to the problem, with differing opinions on where to make the compromise between spatial resolution, temporal resolution, signal-to-noise ratio, and acquisition duration, while the effects of accelerated acquisition schemes (i.e. partial k-space acquisitions) on the quantification accuracy has heretofore remained unknown.
The Solution pursued by our group
We have developed an anthropomorphic phantom device which mimics the male pelvic area, in which precisely and accurately known 'ground truth' CTCs can be generated and presented to the MRI scanner for measurement, thereby allowing for a quantitative assessment of the scanner's ability to accurately measure the contrast agent wash-in and wash-out curve-shapes. This device has allowed us to investigate measurement inaccuracies in these curves-shapes, which may contribute to the wide divergence in DCE PK modelling output parameter values in the published literature, and this information should help in the design of clinical DCE protocols.
Phantom Design and Construction
Prostate mimicking structure, containing two 'measurement chambers'
Photograph of the DCE phantom
Optical calibration set-up: A measurement chamber was connected to the pumps via 11m of tubing, and black dye added to the solution to create an optical signal change which was subsequently measured using a digital camera.
Photo of MRI session, showing the 4 pumps and containers containing the Gd-containing reservoirs
Typical measured STCs with varying dynamic scan times: 1.93, 5s and 14.3s, demonstrating deviations from the ground truth values (solid lines)
- Knight SP, Browne JE, Meaney JF, Smith DS, Fagan AJ, "A novel anthropomorphic flow phantom for the quantitative evaluation of prostate DCE-MRI acquisition techniques", Physics in Medicine and Biology, 61, 7468-7483, 2916, http://dx.doi.org/10.1088/0031-9155/61/20/7466.