Encrypted login | home

Program Information

An MRI-Based Approach to Quantify Radiation-Induced Normal Tissue Injury Applied to Trismus After Head and Neck Cancer Radiotherapy

no image available
M Thor

M Thor1*, N Tyagi1 , V Hazoglou2 , Z Saleh1 , N Riaz3 , N Lee3 , J Deasy1 , (1) Department of Medical Physics, Memorial Sloan Kettering Cancer Center, Nyc, NY, (2) Department of Radiology, Memorial Sloan Kettering Cancer Center, New York, NY, (3) Department of Radiation Oncology, Memorial Sloan Kettering Cancer Center, Nyc, NY


WE-FG-202-7 (Wednesday, August 3, 2016) 1:45 PM - 3:45 PM Room: 202

Purpose: The aim of this study was to investigate if quantitative MRI-derived metrics from four masticatory muscles could explain mouth-opening limitation/trismus following intensity-modulated radiotherapy (IMRT) for head and neck cancer (HNC).

Methods: Fifteen intensity-based MRI metrics were derived from the masseter, lateral and medial pterygoid, and temporalis in T1-weighted scans acquired pre- and post gadolinium injection (T1Pre, T1Post) of 16, of in total 20, patients (8 symptomatic; 8 asymptomatic age/sex/tumor location-matched) treated with IMRT to 70 Gy (median) for HNC in 2005-2009. Trismus was defined as “≥decreased range of motion without impaired eating” (CTCAE.v.3: ≥Grade 1). Trismus status was monitored and MRI scans acquired within 1y post-RT. All MRI-derived metrics were assessed as ΔS=S(T1Pre)-S(T1Post)/S(T1Pre), and were normalized to the corresponding metric of a non-irradiated volume defined in each scan. The T1Pre structures were propagated onto the RT dose distribution, and the max and mean dose (Dmax, Dmean) were extracted. The MRI-derived metrics, Dmax, and Dmean were compared between trismus and non-trismus patients. A two-sided Wilcoxon Signed rank test-based p-value≤0.05 denoted significance.

Results: For all four muscles the population mean of Dmax and Dmean was higher for patients with trismus compared to patients without trismus (ΔDmax=2.3-4.9 Gy; ΔDmean=and 2.0-3.8 Gy). The standard deviation (SD), the variance, and the minimum value (min) of ΔS were significantly (p=0.04-0.05) different between patients with and without trismus with trismus patients having significantly lower SD (population median: -0.53 vs. -0.31) and variance (-2.09 vs. -0.73) of the masseter, and significantly lower min of the medial pterygoid (-0.36 vs. -0.19).

Conclusion: Quantitative MRI-derived metrics of two masticatory muscles were significantly different between patients with and without trismus following RT for HNC. These metrics could serve as image-based biomarkers to better understand the RT-induced etiology behind trismus, but should be further investigated in the complete cohort.

Contact Email: