Purpose: This study aims at analyzing the dependency of normal tissue complication probability (NTCP) modeling on follow-up time and organ dose regarding the endpoint of patient reported (PRO) sticky saliva after radiotherapy.
Methods: A dose of 60Gy treated with IMRT was prescribed to 151 patients with HPV-associated oropharyngeal squamous cell carcinoma. Sticky saliva was assessed based on the PRO-CTCAE score and it was defined as a ≥2 point increase from baseline at 6, 12 and 24 months post-RT. The dose volume histograms (DVH) of the contralateral glands (parotid and submandibular glands combined), contralateral parotid and salivary glands (ipsilateral and contralateral parotid and submandibular glands combined) of each patient were exported. The dose and response data were fitted by the Lyman-Kutcher-Burman (LKB) and Relative Seriality (RS) NTCP models.
Results: Sticky saliva was observed in 31%, 20% and 15% of the patients at 6, 12 and 24 months, respectively. The AUC values were highest (0.7) for the contralateral glands at 12mo post-RT. For the contralateral glands, the values of the D50, m and n parameters of the LKB model were 44.1Gy, 0.43 and 0.65, whereas the values of the D50, γ and s parameters of the RS model were 48.7Gy, 0.74 and 0.92 at 12mo post-RT. Statistically significant Odds Ratios (ORs) of 6.0 (1.3-28.6) and 7.8 (1.7-35.1) were found for the contralateral glands for biological doses (gEUD) < 43Gy and 45Gy for the LKB and RS models, respectively.
Conclusion: The dose-response curves of PRO sticky saliva at 6, 12 and 24 months post-RT could be determined by the LKB and RS NTCP models for different combinations of salivary glands. Keeping the gEUD to the contralateral glands < 43Gy was found to significantly reduce by 6-8 times the risk for patient reported sticky saliva.
NTCP, Dose Response, Radiation Effects
TH- Response Assessment: Modeling: other than machine learning