Session Item

Late radiation induced side effects like swallowing dysfunction are frequently reported by patients treated for head and neck cancers (HNC). The dose to the pharyngeal constrictor muscles (PCM) is considered a prognostic factor and it is recommended to reduce PCM dose during treatment planning as much as possible. However, contouring of PCM are time consuming, the purpose of this study was therefore to evaluate the potential benefit of routine contouring and dose constraining of the PCM in terms of reduction in dose and estimated Normal Tissue Complication Probabilities (NTCP). 

The patient material consisted of 30 patients with HNC in the oral cavity, oropharynx and hypopharynx. Patients was was treated using VMAT with simultaneously integrated boost. Prescribed dose were 68 Gy, 60 Gy and 50 Gy delivered in 6 fractions per week. In 15 of the patients the PCM were contoured during treatment planning according to international consensus guidelines by dedicated RTTs who were certified in OAR contouring after completion of an in-house developed training program (Radiotherapy and Oncology (2015), 117 ; 83-90) . As dose constraint, the mean dose (D_mean) to PCM ≤ 55 Gy was used as a starting point in the plan optimization; PCM dose was further reduced if possible without compromising target dose coverage. Propensity score matching using disease site, bilateral treatment and T-stage category as matching criteria were performed to match the 15 patients where treatment plans were optimized to reduce PCM dose, to the 15 patients where dose optimization of the PCM had not been performed (Table I). For these patients the PCM was contoured retrospectively.  The D_mean to the PCM, the superior, middle and inferior PCM substructures (PCM_sup, PCM_mid and PCM_inf, respectively) were extracted from the dose volume histograms and compared. NTCP for difficulties with swallowing solid food were estimated using a previously published model (Radiotherapy and Oncology (2016), 118; 298-303). Mann-Whitney U-test were used to test for differences between the groups (Optimized vs non-optimized). 

The  D_mean to PCM were less than 55 Gy in all patients in the optimized group and in 10 patients in the non-optimized group. The D_mean was significantly higher in the non-optimized group for the PCM (p = 0.02) and the PCM_sup (p=0.03) (Figure 1a-b/ Table I). No significant group difference was found for PCM_mid and PCM_inf (Table I).  Differences in estimated NTCP was in general higher in the non-optimized group, however the difference was not significant (p= 0.2) (Figure 1c).

Although the dose to PCM was significantly reduced when applying dose constraints during the optimization process, this was not translated into a significant reduction in NTCP. The patient material was however limited and we recommend further analysis in a larger patient cohort where outcome data on swallowing dysfunctions are included.