Synopsis

Assessing response of glioblastoma (GBM) to standard 6-week chemo-radiation at early stages of the treatment allows for changing or adjusting therapy in patients with progressive tumors. Quantitative magnetization transfer (qMT) which probes the properties of the macromolecular content of the tumor and their interactions with free water protons is more sensitive to treatment-induced effects and is capable of differentiating progressors from non-progressors as early as two weeks into the treatment. Moreover, certain qMT parameters (i.e. amount of MT and direct effect of free water pool) were capable of characterizing GBM aggressiveness even before the start of the treatment.

Purpose

Introduction

Materials & Methods

Patients: 19 patients were recruited (REB approved) and were treated for six weeks with concurrent radiation (2Gy/day) and daily Temozolomide. Patients were MRI-scanned 4 times:

1. before treatment (Day₀),

2. After 10 treatment sessions (Day₁₄),

3. After 20 treatment sessions (Day₂₈),

4. One month after last treatment day (Day₇₀).

MRI: was performed on a 3T Philips Achieva scanner. An axial slice passing through the tumor was imaged with TFE (TR/TE=7.78/4.5ms, Matrix=144×144, FOV=20×20cm, slice=3mm).

qMT acquisition: Six MT spectrums were acquired. MT preparation involved 2 & 4 block pulses of 242.5ms each (total duration=970ms & 485ms), and for each duration three RF powers of 1.5/3/5µT were used. MT spectrum covered 250Hz to 100kHz with 14 offsets (with equal log scale separation).

Analysis: All images (MT, T₁, T₂, post Gd-T_1w, FLAIR were co-registered to the first acquired MT image. Tumor ROI was identified at each scan on the post-Gd T_1w. Metrics were calculated voxel-by-voxel and then averaged over tumor ROI.

qMT Modeling: The six MT spectrums were fit to the two-pool MT model proposed by Henkelman⁴ in transient state, and the four model [R, T_2b, RM_0b/R_a, 1/(R_aT_2a)] were calculated where RM_0b/R_a, M_0b, R represent amount of magnetization transfer, macromolecular pool size, and exchange rate.

Results

Tumor progression was determined by an oncologist blinded to the MRI analysis at 3-8 months after the end of the 6-week chemo-radiation treatment and was based on clinically used RANO criteria. Data of three patients were discarded due to incomplete scans or presence of imaging artifacts. Six patients had tumor progression (progressors cohort), the other 10 patients were classified as non-progressors.

Figure1 shows the six MT spectrums as well as the fit to the data for a representative patient. Figure2 shows the qMT parametric maps for a representative patient. R_a^obs and T₂^obs are the observed longitudinal and transverse relaxations of the tissue and were measured independent of the qMT.

Table1 reports the qMT parameters for contralateral normal-appearing white matter (cNAWM) of the patients. Table2 reports qMT model parameters for tumor ROI showing at baseline (Day₀) amount of MT, RM_0b/R_a (p=0.006), and direct effect of the free water pool, 1/(R_aT_2a) (p=0.038), were different between progressors and non-progressors.

Figure 3 plots the ratio of each metric at each time point over its value at Day₀ for RM_0b/R_a, and 1/(R_aT_2a). The ratio represents the change in these parameters.

Discussions

Tables 1,2 show that the amount of MT in GBM tumors was smaller than cNAWM ( RM_0b/R_a,GBM= 1.28±0.44, RM_0b/R_a,cNAWM = 2.89±0.26, p<0.001). There was a statistically significant difference in RM_0b/R_a between progressors and non-progressors before the start of the treatment (Day₀). Non-progressors had a lower amount of MT at baseline and treatment resulted in a large increase in this parameter. However, for progressors, the amount of MT was higher and the treatment was not able to change it significantly (figure3).

Moreover, the direct effect of the free water pool, 1/(R_aT_2a), was lower in non-progressors. Considering R_a is not different between the two cohorts, thus, T_2a was higher in non-progressors, suggesting they had higher water content and lower cellular density. These results demonstrate the ability of qMT in determining GBM tumors that are resistant to standard chemo-radiation even before the start of the treatment.

The ratio of each qMT parameter at each time point over its baseline (Day₀) value represented the treatment-induced changes in the tumor. Figures3 shows the change (between Day₁₄ and Day₀) in the amount of MT (relative RM_0b/R_{a,Non-progressor}=1.34±0.21, relative RM_0b/R_a,Progressor=1.07±0.08 p=0.031), and the direct effect of the free water pool at Day₁₄ (relative 1/(R_aT_2a)_{,Non-progressor}=1.45±0.27, relative 1/(R_aT_2a)_,Progressor=1.14±0.20, p=0.049) were statistically significantly different between progressors and non-progressors. These two parameters reflect the qMT properties of both proton pools in the qMT model ( RM_0b/R_a for the bound macromolecular pool and 1/(R_aT_2a) for the free water pool), showing that chemo-radiation treatment of GBM is affecting both pools.

Conclusions

Acknowledgements

References

1. Ryken, T. C. et al. The role of imaging in the management of progressive glioblastoma: A systematic review and evidence-based clinical practice guideline. J. Neurooncol. 118, 435–460 (2014).

2. Koshy, M. et al. Improved survival time trends for glioblastoma using the SEER 17 population-based registries. J. Neurooncol. 107, 207–212 (2012).

3. Olson, J. J. & Ryken, T. Guidelines for the treatment of newly diagnosed glioblastoma: introduction. Journal of neuro-oncology 89, 255–258 (2008).

4. Henkelman, R. M. et al. Quantitative interpretation of magnetization transfer. Magn. Reson. Med. 29, 759–766 (1993).