Factors associated with dosimetric changes and value of adaptive radiotherapy in patients undergoing radiotherapy for head and neck cancer

Abstract Background: Anatomical changes may occur during radiotherapy (RT) in Head and neck (H&N) cancer patients. These may lead to deviations between the initially planned doses and the delivered doses leading to a risk of organs at risk (OAR) overdose and tumor volume undercoverage. The aim of this study is to identify H&N patients who would benefit from adaptive radiotherapy (ART). Material and Methods: Retrospective data were collected for patients treated with RT for H&N cancer requiring a new dosimetric computed tomography (CT) during treatment. The initially planned doses to OAR and tumor volumes were compared to the doses actually received with and without ART. Predictive criteria for dose deviations greater than 5% were tested. Results: Among 51 patients included in this study, 29 had an overdose to at least one OAR and 13 had a tumor volume undercoverage. Brainstem overdose was associated with cervical volume decrease between the initial dosimetric CT and the 36±2Gy cone beam CT (CBCT) (OR= 11.29, p= 0.027). Homolateral parotid gland (PG) overdose was associated with a decrease of cervical volume (OR= 4.5, p= 0.033) as well as the decrease in high-risk clinical and planned target volumes (HR CTV and PTV) between the initial and the new dosimetric CT [(OR= 3.38, p = 0.04) and (OR= 7.5, p= 0.017 respectively)]. Tumor volume undercoverage was associated with an older age at diagnosis (OR= 6.79, p= 0.022), a small initial macroscopic tumor volume (OR= 9.33, p= 0.006), and a decrease in HR CTV and PTV at the new dosimetric CT (OR= 7.29, p= 0.016; and OR= 7.7, p= 0.006 respectively). With ART, 19 patients (37.3%) showed an improvement in the sparing of at least one OAR and 6 (11.8%) an improvement in tumor volume coverage. Conclusion: Factors such as a decrease in cervical diameters and volume during RT, low initial GTV, high age at diagnosis or a decrease in CTV and PTV HR during RT may help predict possible overdosage to OARs or PTV undercoverage. This might allow a better selection of patients who can benefit from ART.