Dynamic Contrast Enhanced MRI Measurements in Glioma: Comparison Between Two Models
Sameeha Fallatah1, Rolf Jäger1, and Xavier Golay1

1Brain Repair and Rehabilitation, UCL, Institute of Neurology, London, United Kingdom

Synopsis

Dynamic Contrast Enhanced MRI is used to assess the integrity of the blood brain barrier. A major difficulty for the method to be accepted in the clinics is the variety of pharmacokinetic models used and their strong dependence on the underlying assumptions and/or acquisition parameters. Thus the far simpler methods based on signal intensity curve characteristics are the most commonly used approaches in clinical practice. In this study we compare two different pharmacokinetic models, the extended Tofts model and Lawrence & Lee model in patients with primary brain tumours.

Background and Purposes

Dynamic Contrast Enhanced (DCE)-MRI is used to assess the integrity of the blood brain barrier (BBB). The clinical adoption of quantitative DCE-MRI has been slow, largely due to the complexity of the measurements and its interpretation. A major difficulty for the method to be accepted in the clinics is the large variety of pharmacokinetic models used and their strong dependence on the underlying assumptions and/or acquisition parameters. Thus the far simpler methods based on signal intensity curve characteristics are the most commonly used approaches in clinical practice. Nevertheless, the most widely used quantitative model is the extended Tofts model (ETM) . The ETM model allows the quantification of forward volume transfer constant (Ktrans), extravascular extracellular space volume (Ve), blood plasma volume (Vp) and the reverse vascular transfer constant (Kep), based on a simple two-pool model. These parameters reflect permeability as well as perfusion. On the other hand, the Lawrence & Lee model (L&L) accounts for the small changes of concentration in contrast agent over time, thus allows for flow (F), extraction fraction (E) and mean capillary transit time (Tc) quantification. We aim in this study to compare the values of Ktrans, ve, Kep and ve obtained by both models in a series of patients with primary brain tumours.

Methods

Subjects: 27 patients, age 45 ±15.4 years, 15 males and 12 females. 6 Astrocytoma WHO II, 5 Oligodendroglioma WHO II, 4 Oligoastrocytoma WHO II, 4 anaplastic Oligodendroglioma WHO III, 2 Anaplastic oligoastrocytoma WHO III, 3 anaplastic astrocytoma WHO III, 3 GBM. All underwent DCE-MRI scanning as part of their standard clinical care. All data were post-processed retrospectively as part of an approved audit by our institution.

Post-processing: DCE data were post-processed twice with Olea Sphere™ software using both the ETM and L&L models. Four parametric maps were produced Ktrans, ve, vp and Kep with each model.

Image analysis: A 20-30 mm2 ROI with the highest Ktrans value was used then propagated to other DCE-MRI maps for both models to extract the mean parameters (Ktrans, ve, vp and Kep). Paired t-tests and Pearson's correlation coefficient are used to study the relationship between all parameters given by both DCE models.

Results

DCE parameter values produced by the two DCE models are listed in table 1. Although the measurements produced by the 2 models are highly correlated (table 2), the differences were generally very large, with the exception of Ktrans in all tumours, or for low-grade astrocytomas (LGA) in general. Parameters produced by the L&L model are generallt higher than those produced by the ETM model, with mean differences 0.02, 0.25, 0.31 and 0.63 for Ktrans, Vp, Ve and Kep respectively.

Discussion and conclusion

The model used can influence the resulting quantitative permeability parameters, especially in lesions with high-level of leakage (HGA, HGO, LGO). However both models have a strong linear relationship for all parameters, which means that the parameters from each model give similar information, but on a different scale. The reason for the lack of difference between models for LGA is likely due to the low leakage rate in these lesions (Ktrans =0.001-0.002)

Acknowledgements

No acknowledgement found.

References

1. Patlak CS etal. Journal of Cerebral Blood Flow and Metabolism 1985;5(4):584–90.

2. Tofts PS etal. J Magn Reson Imaging 1999;10(3):223–32.

3. Lawrence, KS Lee T-Y. Journal of Cerebral Blood Flow and Metabolism 1998;18(12):1378–85.

4. Buckley DL. Magn Reson Med 2002;47(3):601–6.

5. Cramer SP etal . J Cereb Blood Flow Metab 2014;34(10):1655–65.

Figures

Figure1: DCE maps for anaplastic oligodendroglioma using both models. ETM, extended Tofts model (top row) and L&L, Lawrence & Lee (bottom row)

Table1: Mean± standard deviation of DCE parameters, for each tumour group, produced by L&L, Lawrence&Lee and ETM, extended Tofts model. HGA, anaplastic astrocytoma/GBM; LGA, low-grade astrocytomas; HGO, anaplastic oligodendro /oligoastrocytoma; LGO, low-grade oligodendro /oligoastrocytoma.

Table2: P values to discribe the differences between L&L, and ETM for each tumour subtype; r is the correlations coefficient and the mean difference between the 2 models.



Proc. Intl. Soc. Mag. Reson. Med. 24 (2016)
1932