Abstract
BACKGROUND: Thoracic lymph nodes are critical targets in radiotherapy for locally advanced non-small cell lung cancer (LA-NSCLC), but their accurate localization is complicated by interfractional shifts. PURPOSE: To evaluate the interfractional carina and hilum shifts as surrogates for thoracic lymph node areas during conventional fractionated radiotherapy for LA-NSCLC and to explore baseline characteristics of patients related to these shifts. METHODS: For 23 patients, the carina and hilum nearest the primary tumor were marked on daily cone-beam computed tomography (CBCT) images after vertebrae-based registration. The interfractional shifts of these two points were determined based on comparison of the first and subsequent CBCT scans. The patients were grouped using dynamic time warping clustering, and their baseline characteristics were compared. Population margins were calculated using the van Herk formula. Shift directionality was analyzed using principal component analysis. RESULTS: The proportion of shifts of >5 mm in all CBCT scans was 10.9% for the carina and 22.8% for the hilum. The patients were grouped into small-shift, medium-shift, and large-shift groups based on their hilum shifts. The large-shift group had the largest shifts, with a median (interquartile range) of 3.9 mm (2.4-5.2) for the carina and 6.3 mm (4.4-10.1) for the hilum. The gross tumor volume (GTV) within 2 cm of the central airways (proximal GTV) was significantly different among groups (p < 0.01). Greater proximal GTVs were correlated with large median shifts of each patient (correlation coefficients: carina, 0.57; hilum, 0.76) and showed a tendency toward unidirectional hilum shift. Population margins reached 4.7-7.5 mm for the carina and 6.9-13.1 mm for the hilum in the highest proximal GTV tertile (median 30.6 cm(3)). CONCLUSIONS: Variations were observed in the shifts across patients and lymph node positions. Greater proximal GTVs were correlated with large and directional hilum shifts, indicating the potential benefits of tailoring adaptive radiotherapy strategies.