Quantitative DCE-MRI for differentiating high grade glioma recurrence from treatment-related changes:  Effect of T1 mapping method
Greg O. Cron1,2,3, Beckie Manouchehri4, Andrew Boivin3, Nader Zakhari1,3, Brandon Zanette5,6, Gerard H. Jansen1,2,3, John Woulfe1,2,3, Rebecca E. Thornhill1,2,3, Andreas Greiser7, Ian G. Cameron1,2,3,4, and Thanh B. Nguyen1,2,3

1The Ottawa Hospital, Ottawa, ON, Canada, 2Ottawa Hospital Research Institute, Ottawa, ON, Canada, 3University of Ottawa, Ottawa, ON, Canada, 4Carleton University, Ottawa, ON, Canada, 5University of Toronto, Toronto, ON, Canada, 6Peter Gilgan Centre for Research and Learning, The Hospital for Sick Children, Toronto, ON, Canada, 7Siemens Healthcare GmbH, Erlangen, Germany

Synopsis

After a patient has received treatment for a high grade glioma, a new enhancing lesion presents a common diagnostic dilemma: Malignant tumor recurrence and benign treatment-related changes (TRC) appear similar on conventional MRI. MRI tracer kinetic studies may help distinguish recurrence from TRC. We investigated whether Ktrans measurements using quantitative DCE-MRI can accurately diagnose recurrence. We also studied the effect of the DCE T1 mapping method (VFA versus LL). Ktrans values in recurrent tumor were higher than in TRC, providing sensitivity of 69-77%, specificity of 100%. The choice of T1 mapping had little effect on diagnostic accuracy.

PURPOSE

After a patient has received surgery and chemoradiation treatment for a high grade glioma, the appearance of a new enhancing lesion on conventional contrast-enhanced MRI presents a common diagnostic dilemma: Malignant tumor recurrence and benign treatment-related changes (TRC) appear similar (1-3). Some investigators have found that MRI tracer kinetic studies may help distinguish high grade glioma recurrence from TRC (3-10). For these kinetic studies, it can be argued that the best accuracy and reproducibility should be achieved using quantitative dynamic contrast-enhanced (DCE) MRI, where T1 mapping is employed to help convert dynamic changes in MRI signal into concentration-vs-time curves (11-13).

Regarding the T1 mapping method used for DCE-MRI, there are a large number from which to choose. One common technique is uncorrected variable flip angle (VFA), which employs a gradient echo sequence with different flip angles, and no correction for flip angle error (14-15). Another technique is "Look-Locker" (LL), which uses an inversion pulse followed by rapid sampling of the longitudinal magnetization (16-18). T1 mapping methods which use inversion pulses are more time consuming, but usually more accurate, than variable flip angle methods (19-20).

It has been found previously that the choice of T1 mapping method can potentially affect quantitative DCE-MRI (20-21). We therefore sought to investigate the effect of the T1 mapping method (VFA versus LL) on the ability of quantitative DCE-MRI to distinguish high grade glioma recurrence from TRC.

METHODS

This prospective study included 22 patients with high grade glioma who developed a new enhancing lesion on post-treatment MR. They were classified as tumor recurrence or TRC based on histopathological analysis or clinical/imaging follow-up. Each patient underwent an MRI examination, including the DCE-MRI, on a clinical 3T scanner (Magnetom Trio, Siemens Healthcare).

For DCE-MRI, a baseline T1 map was obtained with a modified LL method ("MOLLI", prototype WIP448). Next, another baseline T1 map was obtained with uncorrected VFA ("MapIt" sequence). 3D 18-slice DCE-MRI was then performed, using signal phase to compute a vascular input function (22-23). Immediately after DCE-MRI, VFA and LL T1 mapping were repeated. Using the VFA or LL T1 maps, the "Bookend Method" was used to convert tissue signal-vs-time to concentration-vs-time voxel-by-voxel (20-23).

Each of the two DCE-MRI datasets (VFA, LL) was analyzed using the Extended Tofts Model (ETM) in a commercial software package (OLEA Sphere, ©Olea Medical, version 2.2 SP3) to obtain Ktrans (volume transfer constant) and Vp (plasma volume) voxel-by-voxel. For each patient, a whole-tumor region of interest was drawn on a central tumor slice, from which median Ktrans_VFA, Ktrans_LL, Vp_VFA, and Vp_LL were computed. Mann-Whitney, Bland-Altman, and ROC analyses were performed with MedCalc® Version 13.1.1.0.

RESULTS

Median Ktrans values in recurrent tumor were higher than in TRC (P=0.004, Fig 1). Ktrans_VFA and Ktrans_LL in TRC were 0.014 (0.008-0.017) and 0.011 (0.009-0.015) min-1, respectively [med (95% CI)], whereas Ktrans_VFA and Ktrans_LL in recurrent tumor were 0.037 (0.021-0.068) and 0.031 (0.021-0.047) min-1. Bland-Altman analysis showed that Ktrans_VFA and Ktrans_LL values were similar, with Ktrans_VFA about 0.01 min-1 higher on average (Fig 2).

For diagnosing tumor recurrence, Ktrans had low sensitivity (69-77%) but high specificity (100%) (Table 1). There was no statistically significant difference in diagnostic accuracy between Ktrans_VFA and Ktrans_LL (Fig 3). The diagnostic accuracy of both Vp_VFA and Vp_LL was poor (P>0.4 for difference between AUC and 0.5).

DISCUSSION

The results of this study show that DCE-derived Ktrans (but not Vp) is useful for distinguishing high grade glioma recurrence from TRC. A similar Ktrans result was reported by Bisdas et al (8). The negative result for Vp is somewhat surprising, given that some investigators have found DSC-derived cerebral blood volume (CBV, which should be an analogue of Vp) (3-5) and DCE-derived CBV (7) to have diagnostic value for this application. The latter DCE study (7) used a Patlak analysis instead of ETM, and did not find other parameters (e.g. CBF) useful. Moreover, in patients with newly diagnosed astrocytomas, DSC-derived CBV and DCE-derived Vp do not necessarily correlate well (22), reflecting the different nature of the two measurements.

The choice of T1 mapping method (VFA or LL) for this study had no effect on diagnostic accuracy, and the Ktrans values of the two methods were similar. This is perhaps not too surprising, since these types of applications are often dominated by biological variation. The LL method would likely provide better inter-scanner and inter-institutional standardization, however (20).

CONCLUSION

When using quantitative DCE-MRI to distinguish high grade glioma recurrence from treatment-related changes, Ktrans is more useful than Vp. Additionally, the choice of T1 mapping has little effect on diagnostic accuracy.

Acknowledgements

This work is supported by research grants from the Brain Tumour Foundation of Canada and Cedars Cancer Foundation.

References

1. Shah, Ashish H., et al. "Discriminating radiation necrosis from tumor progression in gliomas: a systematic review what is the best imaging modality?." Journal of neuro-oncology 112.2 (2013): 141-152.

2. Caroline, I., and M. A. Rosenthal. "Imaging modalities in high-grade gliomas: pseudoprogression, recurrence, or necrosis?." Journal of Clinical Neuroscience19.5 (2012): 633-637.

3. Parvez, Kashif, Aatif Parvez, and Gelareh Zadeh. "The diagnosis and treatment of pseudoprogression, radiation necrosis and brain tumor recurrence."International journal of molecular sciences 15.7 (2014): 11832-11846.

4. Fatterpekar, Girish M., et al. "Treatment-related change versus tumor recurrence in high-grade gliomas: a diagnostic conundrum—use of dynamic susceptibility contrast-enhanced (DSC) perfusion MRI." American Journal of Roentgenology 198.1 (2012): 19-26.

5. Barajas Jr, Ramon F., et al. "Differentiation of recurrent glioblastoma multiforme from radiation necrosis after external beam radiation therapy with dynamic susceptibility-weighted contrast-enhanced perfusion MR imaging 1." Radiology 253.2 (2009): 486-496.

6. Kim, Dong Hyeon, et al. "Differentiation of True Recurrence from Delayed Radiation Therapy-related Changes in Primary Brain Tumors Using Diffusion-weighted Imaging, Dynamic Susceptibility Contrast Perfusion Imaging, and Susceptibility-weighted Imaging." Journal of the Korean Society of Magnetic Resonance in Medicine 18.2 (2014): 120-132.

7. Larsen, Vibeke A., et al. "Evaluation of dynamic contrast-enhanced T1-weighted perfusion MRI in the differentiation of tumor recurrence from radiation necrosis." Neuroradiology 55.3 (2013): 361-369.

8. Bisdas, Sotirios, et al. "Distinguishing recurrent high-grade gliomas from radiation injury: a pilot study using dynamic contrast-enhanced MR imaging." Academic radiology 18.5 (2011): 575-583.

9. Shin, K. E., et al. "DCE and DSC MR perfusion imaging in the differentiation of recurrent tumour from treatment-related changes in patients with glioma." Clinical radiology 69.6 (2014): e264-e272.

10. Narang, Jayant, et al. "Differentiating treatment-induced necrosis from recurrent/progressive brain tumor using nonmodel-based semiquantitative indices derived from dynamic contrast-enhanced T1-weighted MR perfusion." Neuro-oncology (2011): nor075.

11. Paldino, Michael J., and Daniel P. Barboriak. "Fundamentals of quantitative dynamic contrast-enhanced MR imaging." Magnetic resonance imaging clinics of North America 17.2 (2009): 277-289.

12. Rijpkema, Mark, et al. "Method for quantitative mapping of dynamic MRI contrast agent uptake in human tumors." Journal of Magnetic Resonance Imaging 14.4 (2001): 457-463.

13. Tietze, Anna, Kim Mouridsen, and Irene Klærke Mikkelsen. "The impact of reliable prebolus T1 measurements or a fixed T1 value in the assessment of glioma patients with dynamic contrast enhancing MRI." Neuroradiology 57.6 (2015): 561-572.

14. Homer J, Reevers MS. Driven-Equilibrium Single-Pulse Observation of T1 A Reevaluation of a Rapid “New” Method for Determining NMR Spin-Lattice Relaxation Times. J. Magn. Reson. 1985;63:287–297.

15. Brookes JA, Redpath TW, Gilbert FJ, Murray AD, Staff RT. Accuracy of T1 measurement in dynamic contrast-enhanced breast MRI using two- and three-dimensional variable flip angle fast low-angle shot. J. Magn. Reson. Imaging 1999;9:163–71.

16. Look DC, Locker DR. Time Saving in Measurement of NMR and EPR Relaxation Times. Rev. Sci. Instrum. 1970;41:250.23.

17. Zhang YT, Yeung HN, Carson PL, Ellis JH. Experimental Analysis of T1 imaging with a Single-Scan, Multiple-Point, Inversion-Recovery Technique. Magn. Reson. Med. 1992;25:337–343.24.

18. Messroghli DR, Radjenovic A, Kozerke S, Higgins DM, Sivananthan MU, Ridgway JP. Modified Look-Locker inversion recovery (MOLLI) for high-resolution T1 mapping of the heart. Magn. Reson. Med. 2004;52:141–6

19. Siversson, Carl, et al. "Repeatability of T1-quantification in dGEMRIC for three different acquisition techniques: Two-dimensional inversion recovery, three-dimensional look locker, and three-dimensional variable flip angle." Journal of Magnetic Resonance Imaging 31.5 (2010): 1203-1209.L

20. Zanette, Brandon, et al. "Accuracy of quantitative 3D DCE-MRI using variable flip angle T1 mapping, B1 correction, and the bookend method." Proceedings of the 21st Annual Meeting of the International Society of Magnetic Resonance Imaging in Medicine. 2013.

21. Cron, Greg O., Giles Santyr, and Frederick Kelcz. "Accurate and rapid quantitative dynamic contrast-enhanced breast MR imaging using spoiled gradient-recalled echoes and bookend T 1 measurements." Magnetic resonance in medicine 42.4 (1999): 746-753.

22. Nguyen, T. B., et al. "Comparison of the Diagnostic Accuracy of DSC-and Dynamic Contrast-Enhanced MRI in the Preoperative Grading of Astrocytomas." American Journal of Neuroradiology (2015).

23. Nguyen, T. B., et al. "Preoperative Prognostic Value of Dynamic Contrast-Enhanced MRI–Derived Contrast Transfer Coefficient and Plasma Volume in Patients with Cerebral Gliomas." American Journal of Neuroradiology 36.1 (2015): 63-69.

Figures

Fig 1: Two example patients. Top row: Lesion caused by treatment-related changes (TRC). Bottom row: Tumor recurrence. Left column: Conventional contrast-enhanced MRI. Middle column: Ktrans maps from DCE-MRI using VFA T1 mapping. Right column: Ktrans maps from DCE-MRI using Look-Locker T1 mapping. Number below each map shows median Ktrans value in the region of interest, which is delineated by a green outline.

Fig 2: Bland-Altman plot for Ktrans values in all patients, VFA versus LL. The two methods returned similar values, although VFA Ktrans values tended to be higher.


Fig 3: Receiver operating characteristic (ROC) curves showing the ability of Ktrans to diagnose recurrent tumor. Blue line = VFA method (AUC ± SE = 0.87 ± 0.08); red dotted line = Look-Locker method (AUC ± SE = 0.87 ± 0.09). An AUC between 0.8 and 0.9 is considered a "good" diagnostic test. There was no statistically significant difference between the two ROC curves (P=1.0).


Table 1: ROC characteristics of Ktrans VFA and Ktrans Look-Locker, showing the ability of Ktrans to diagnose recurrent tumor. Both methods showed low sensitivity but excellent specificity. aDelong et al.; bBinomial exact.



Proc. Intl. Soc. Mag. Reson. Med. 24 (2016)
4354