19 patients (18 male, 1 female; median age, 58 years; range, 46 – 67 years) with loco-regionally advanced HNSCC treated with concomitant CRT were included in this study.
Pretreatment MR images (1 to 7 days before CRT start) and MR images obtained 2.5 to 3 months after the end of treatment were analysed retrospectively. MR imaging was performed on 3T with a neck array coil.
Morphologic images including T2-weighted sequences with short tau inversion recovery (STIR) (TR/TE 5010/71 ms, TI 170 ms, slice thickness of 3 mm, a 0.3 mm gap between slices and a field of view (FOV) of 18x18 cm and 256x256 matrix) and post-contrast T1 sequences were acquired in transversal plane from the level of the skull base to the level of the aortic arch.
DWI sequences were acquired in transversal plane with pulsed spin-echo echo-planar STIR sequences (TR/TE 3600/86 ms, slice thickness of 5 mm, gap between slices 1.5 mm, FOV 230 cm2) at five b values (b=0, 100, 200, 500 and 1000 s/mm2).
For the post-contrast T1 volumetric interpolated breath-hold examination (VIBE) sequences ((TR/TE 3.26/1.26 ms, voxel size 1.1x0.9x1.5 mm with 4 mm averages, receiver bandwidth 640 Hz/px, matrix size 218x288 and FOV 250 cm2), the patients received 0.1 mmol/kg of i.v. gadolinium-based contrast agent gadobutrol, followed by 20 ml of saline, both at 3.5 ml/s flow rate, using a power injector.
Post-processing of all the images was performed at a workstation by commercially available software. Before data analysis, motion correction algorithm was applied.
Multiparametric IVIM-DWI (ADC; diffusion coefficient, D; perfusion-related diffusion coefficient, D*; perfusion fraction, f) maps of each primary tumour were generated.
The signal variation with increasing b values was modelled by using the following bi-exponential function by Le Bihan [2]:
Sb/S0=(1-f)×e(-bD)+f×e(-bD*)
This is function of signal variation, where Sb is the signal intensity with diffusion weighting b, S0 is the signal intensity for b-value of 0 s/mm2, f is the perfusion fraction, D is the diffusion coefficient and D* is the perfusion-related diffusion coefficient.
Apparent diffusion coefficient (ADC) maps were computed by a single exponential fit using the DWI signal intensity b-value curves.
Three regions of interests were placed manually in the primary tumour and the average value from all parameters was calculated.
The differences between pretreatment ADC and IVIM parameters of patients with complete response (CR) and partial response (PR) to CRT were compared with Student's t-test. The role of pretreatment IVIM parameters for treatment response was determined using receiver operating characteristic (ROC) curves. Correlations between IVIM parameters and ADC values were evaluated with Pearson's correlation coefficient.