External validation of a multifactorial normal tissue complication probability model for tube feeding dependence at 6 months after definitive radiotherapy for head and neck cancer

Background and purpose: The purpose of this study was to externally validate a previously published normal tissue complication probability (NTCP) model for tube feeding dependence at 6 months (TUBEM6) after completion of (chemo) radiotherapy.

Materials and methods: This study evaluated 122 head and neck cancer patients treated by definitive (chemo) radiotherapy. The closed testing procedure was used to select the appropriate method for updating the NTCP model. In this procedure, the likelihood ratio test was used to compare the updated model against the original model.

Results: Mean predicted NTCP was 12.2% (95% CI: 9.9%-14.5%) when using the original NTCP model for TUBEM6. TUBEM6 at our institute was 5.7% (95% CI: 1.8-9.6%) for the 122 patients evaluated. The test for the model revision against the original NTCP model was statistically significant (p = 0.032). The test for the model revision against the model adjusting intercept only was not statistically significant (p = 0.240). According to the closed testing procedure, the model required adjusting the intercept only.

Conclusions: TUBEM6 at our institute was lower than that predicted by the original NTCP model. The closed testing procedure indicated that only an adjustment of the intercept was needed indicating the importance of external validation. (C) 2018 Elsevier B.V. All rights reserved.