Original Article
The tumor shape changes of nasopharyngeal cancer during chemoradiotherapy: the estimated margin to cover the geometrical variation
Abstract
Background: Considerable geometrical change occurs during chemoradiotherapy (CRT) course of nasopharyngeal carcinoma (NPC). This aim of this study was to quantify the volumetric and surface variability of the target volumes (TV) and to estimate the expanded margin to maintain acceptable geometrical coverage.
Methods: Twenty patients with locally advanced nasopharyngeal cancer underwent one planning CT (pCT) and six weekly repeated CT (rCT) scans during the treatment course of definitive CRT. The TV included the gross tumor volume (GTV) of the primary tumor, large (shortest diameter >3.0 cm) and small (diameter >1 cm and ≤3 cm) positive neck lymph nodes, and low-risk clinical target volume (CTV_Lr) that were delineated manually on the pCT and each rCT. When comparing TV in pCT (V_pCT) and TV in rCT (V_rCT), the overlapping index (OI), Dice similarity coefficient (DSC), shortest perpendicular distance (SPD), and overall standard deviation (overall SD) were calculated to present the geometric changes. An isotropical margin was expanded outward around CTV_Lr in pCT to establish the mimic planning target volume (PTV). An OI ≥0.95 was defined as acceptable geometrical coverage.
Results: For all TV, DSCs decreased, and the SPDs and overall SD increased with the increasing number of fractions delivered. The DSCs of all gross TV were <70% after the third week. The mean SPDs were 1.5–2.5 mm in the first week and 5.2–6.2 mm in the last week. The OI and DSC in concurrent CRT were smaller than those in the sequential therapy; and similarly the SPD and overall SD in the concurrent therapy were larger than those in the sequential one. To maintain >95% geometrical coverage, a 2 mm additional margin could maintain the coverage throughout the treatment course and a 1 mm margin could maintain the desired coverage if there is an adaptive re-planning no later than the third week of the treatment course.
Conclusions: Both volumetric coverage and surface of the tumour underwent the progressive changes during the treatment course of CRT. One to two mm as the expanded margin to establish the PTV is required to maintain >95% geometrical coverage.
Methods: Twenty patients with locally advanced nasopharyngeal cancer underwent one planning CT (pCT) and six weekly repeated CT (rCT) scans during the treatment course of definitive CRT. The TV included the gross tumor volume (GTV) of the primary tumor, large (shortest diameter >3.0 cm) and small (diameter >1 cm and ≤3 cm) positive neck lymph nodes, and low-risk clinical target volume (CTV_Lr) that were delineated manually on the pCT and each rCT. When comparing TV in pCT (V_pCT) and TV in rCT (V_rCT), the overlapping index (OI), Dice similarity coefficient (DSC), shortest perpendicular distance (SPD), and overall standard deviation (overall SD) were calculated to present the geometric changes. An isotropical margin was expanded outward around CTV_Lr in pCT to establish the mimic planning target volume (PTV). An OI ≥0.95 was defined as acceptable geometrical coverage.
Results: For all TV, DSCs decreased, and the SPDs and overall SD increased with the increasing number of fractions delivered. The DSCs of all gross TV were <70% after the third week. The mean SPDs were 1.5–2.5 mm in the first week and 5.2–6.2 mm in the last week. The OI and DSC in concurrent CRT were smaller than those in the sequential therapy; and similarly the SPD and overall SD in the concurrent therapy were larger than those in the sequential one. To maintain >95% geometrical coverage, a 2 mm additional margin could maintain the coverage throughout the treatment course and a 1 mm margin could maintain the desired coverage if there is an adaptive re-planning no later than the third week of the treatment course.
Conclusions: Both volumetric coverage and surface of the tumour underwent the progressive changes during the treatment course of CRT. One to two mm as the expanded margin to establish the PTV is required to maintain >95% geometrical coverage.
