Click here to

Session: Multi-Disciplinary: Imaging for Treatment Planning [Return to Session]

Squamous Cell Tumor Characterization in Head and Neck Cancer Patients Using Dual-Energy Computed Tomography

É Delisle1*, H Bahig2, 3, 4, H Bouchard1, 2, 3, M Simard1, C Larose2, E Filion2, P Nguyen-tân2, S Bedwani1, 2, 3, (1) Physics department, Université de Montréal, Montreal, QC, CA, (2) Radiation Oncology Department, Centre Hospitalier de Université de Montréal, Montreal, QC, CA, (3) Centre de recherche du Centre hospitalier de Université de Montréal, Montreal, QC, CA, (4) Radiation Oncology Department, University of Texas MD Anderson Cancer Center, Houston, TX, USA

Presentations

SU-IePD-TRACK 3-6 (Sunday, 7/25/2021) 5:30 PM - 6:00 PM [Eastern Time (GMT-4)]

Purpose: To characterize contrast agent concentrations in head and neck squamous cell cancers (SCC) as a function of tumor subsite using contrast enhanced dual-energy computed tomography (CE-DECT) pre-treatment scans.

Methods: Patients with larynx, hypopharynx and oropharynx SCC that underwent treatment planning with CE-DECT of the neck were studied. Primary and nodal gross tumor volumes (GTVp, GTVn) were contoured and their CE-DECT data were analysed. A two-material decomposition method was performed to extract iodine concentration (IC) maps. The IC distributions were grouped according to GTV (GTVp, GTVn) and tumor subsite (larynx-hypopharynx, oropharynx). The oropharynx and larynx-hypopharynx IC distributions were compared for both GTVp and GTVn subgroups. Both a Student's t-test and a Welsh t-test were realized to compare the GTVp and GTVn IC distribution's average 90th percentile, mean, variance, skewness and kurtosis. A Welsh t-test was used when the variances of the sample means were unequal.

Results: 65 patients of which 57 had a GTVp and 30 a GTVn were analysed. 35%/65% of GTVp and 73%/27% of GTVn were in the oropharynx/larynx-hypopharynx subsites. There was a significant difference in the average 90th percentile, mean, variance, skewness and kurtosis of GTVp ICs between the larynx-hypopharynx (p90=9.97, M=2.00, VAR=8.80, SKEW=2.15, KURT=3.35) and oropharynx (p90=7.85, M=1.53, VAR=7.78, SKEW=2.58, KURT=8.20) subgroups: p_p90=.0002, p_M=.007, p_VAR=.001, p_SKEW=.004, p_KURT=.002. There was no significant difference in the average 90th percentile, mean, variance, skewness and kurtosis of GTVn ICs between the larynx-hypopharynx and oropharynx subgroups. These preliminary results suggest that IC maps have a better clinical resolving power for GTVp than for GTVn.

Conclusion: This work implies that IC maps could be used as a biomarker to gain further insight into head and neck tumors. This information may be useful when investigating the prediction of locoregional recurrence using tumor IC distributions.

Funding Support, Disclosures, and Conflict of Interest: This work is funded by Varian Medical Systems.

ePosters

    Keywords

    Dual-energy Imaging, CT, Contrast Agent

    Taxonomy

    IM- CT: Dual Energy and Spectral

    Contact Email

    Share: