Diffusion weighted imaging in submandibular gland sparing helical tomotherapy for nasopharyngeal carcinoma.

Abstract

To verify clinical significance of submandibular gland (SMG)-sparing during helical tomotherapy (HT) for nasopharyngeal carcinoma (NPC) from the perspective of imaging by using diffusion weighted imaging (DWI). In this prospective study, 60 NPC patients scheduled for radical SMG-sparing HT were enrolled. All patients underwent DWI examinations prior to HT (pre-HT) and 1, 3, 6, 9, 12 months post HT. Mean apparent diffusion coefficient (ADC) values of bilateral parotid glands (PGs) and submandibular glands (SMGs) were measured. Differences of ADC and changes of ADC pre and pro HT (ΔADC) among SMG-spared, SMG-unspared and PGs were compared and the associations betweenΔADC and variations of patient-rated xerostomia questionnaire summary scores (XQ-sum) were further tested. ADC_post-HT and ΔADC_post-HT of SMG-spared were both much lower than of SMG-unspared and a strong dose-response relationship was detected between mean radiation dose and ΔADC of SMGs. Dynamic change trends of PGs, SMG-spared and SMG-unspared were similar, with initial increase at 1 m-post-HT followed by little change at 3 m-post-HT and then gradual decrease over time. But for SMG-unspared, there was no obvious change of ADC from 6 m-post-HT to 12 m-post-HT. The dynamic change trend of XQ-sum was nearly in line with that of ADC on the whole. And a positive correlation between mean ΔADC_1m-post-HT of bilateral SMGs and variation of XQ-sum_1m-post-HT in patients with bSMG-unspared were found (r = 0.693, P < 0.001). Multivariate stepwise regression analysis showed that whether spared SMG or not was the only independent predictor correlated to XQ-sum_post-HT at each follow-up timepoint. SMG-sparing technique could significantly improve subjective xerostomia post HT in NPC patients from the perspective of imaging.