7th ICHNO Abstract book

page 8 7 th ICHNO Conference International Conference on innovative approaches in Head and Neck Oncology 14 – 16 March 2019 Barcelona, Spain __________________________________________________________________________________________ 7th ICHNO

OC-008 Fast tumor regression detected by weekly MRI allows for adaptive (chemo) radiation for Head and Neck cancer C.Terhaard 1 , B. Peltenburg 1 , R. De Bree 2 , N. Raaijmakers 1 , J.W. Dankbaar 3 , F. Pameijer 4 , M. Philippens 1 1 UMC Utrecht, Radiation Oncology Department, Utrecht, The Netherlands; 2 UMC Utrecht, Department of Head and Neck surgical Oncology, Utrecht, The Netherlands; 3 UMC Utrecht, Radiology Department, Utrecht, The Netherlands; 4 UMC Utrecht, Radiology Department, Utrecht, The Netherlands Purpose or Objective The effect of (chemo)radiotherapy on head and neck squamous cell carcinoma can be visualized using MRI. Adaptation of radiotherapy treatment plans might be possible based on changes in tumor volume during treatment, using MRI guided treatment. Furthermore, different patterns of volume changes might be an indicator of radiosensitivity and could possibly predict treatment failure. Therefore the objective of this study is to follow treatment response with T2 weighted images of head and neck cancer and to adapt based on this tumor regression. Material and Methods Twenty patients with stage II, III or IV oropharyngeal, laryngeal or hypopharyngeal squamous cell carcinoma were included. Two patients had HPV positive tumors. Treatment consisted of radiotherapy, with or without concurrent chemotherapy. All patients underwent MRI prior to and in week 2, 3, 4 and 5 of the radiotherapy treatment.Imaging was obtained with the patient positioned in the radiotherapy mask. Tumor delineation was performed on T2 weighted MR images with fat- suppression (mDIXON) for the baseline and each subsequent images. Relative volume changes compared to the baseline were determined. A follow up of at least 3 months was available for all patients. Results During (chemo)radiotherapy tumors generally reduced in size with each passing week. On average the tumors were only 50% of their original size at the end of the third week of treatment. At the end of the fifth week only on average 20% of the original tumor volume was visible on T2 weighted images (fig 1). In one patient, tumor visibly increased in size from the third week onward. This was the only patient with a local recurrence within 3 months after treatment (fig 2). However, in some cases, tumors were increasingly in time harder to differentiate from nonmalignant tissues in the treatment area. Fig 1. Change in tumor volume of all patients (black). Red line shows the average tumor change across all patients. Purple line identifies the patient with an increasing tumor volume starting in the third week and an eventual local recurrence. Fig 2. T2 weighted MRI images of the patient with a local recurrence within 3 months after treatment. The patient was diagnosed with a T4aN3bM0 hypopharyngeal carcinoma. A) Pretreament MRI with the tumor visible at the right side of the hypopharynx. B) Week 2 of radiotherapy. C) Week 3 of radiotherapy. D) Week 4 of radiotherapy. E) Week 5 of radiotherapy. F) T2 weighted MRI a time of local recurrence (3 months after radiotherapy).

Cancer Centre, Department of Biostatistics, Toronto, Canada; 5 Princess Margaret Cancer Centre, Division of Medical Oncology, Toronto, Canada; 6 Princess Margaret Cancer Centre, Otolaryngology - Head & Neck Surgery, Toronto, Canada Purpose or Objective The 8 th edition TNM (TNM-8) cT1-2N1 HPV+ oropharyngeal cancer (OPC) is generally considered early stage with excellent prognosis. This study aims to identify adverse radiologic nodal features that may portend poorer prognosis in a subset of this population. Material and Methods All TNM-8 cT1-T2N1 HPV+ OPC patients (pts) treated with definitive IMRT in 2008-2015 were included. Tumor HPV status was ascertained by p16 staining. Pre-IMRT CT/MR were reviewed by a designated head and neck radiologist blinded to treatment outcomes. Number and level of radiologically involved lymph node (LN), extranodal extension (rENE: LN with an unequivocal ‘ill-defined’ border), retropharyngeal LN (RPLN), and lower neck (Level 4 or 5b) LNs were recorded. Inter- and intra-rater (after 3-month interval) concordance for rENE were assessed in a randomly selected cohort (n=45). Disease-free survival (DFS), locoregional control (LRC), and distant control (DC) were compared between those with (rENE+) vs without (rENE–). Univariable (UVA) and multivariable analysis (MVA) with step-wise modal selection identified prognostic factors for DFS. Results We excluded 28/308 pts due to unavailable pre-IMRT CT/MR, leaving 280 eligible for analysis [rENE+: 45 (16%); rENE–: 235 (84%)]. The Kappa scores were 0.89 (95% CI: 0.74-1.00) for inter-rater and 0.82 (0.62-1.00) for intra- rater concordance on rENE. Both rENE+ and rENE– cohorts had similar distribution in age (p=0.88), gender (p=0.26), and T category (p=0.40), and presence of RPLN (p=0.07). Proportion of systemic agent usage was similar between rENE+ and rENE respectively (chemotherapy: 29 vs 136; EGFR inhibitor: 1 vs 29) (p=0.13). The rENE+ cohort had a higher number of LNs per pt [median: 6 (1-20) vs 2 (1-15), p<0.001] and was more likely to have necrotic LNs [33 (73%) vs 132 (56%), p=0.046]. Median follow-up was 4.8 years. Although LRC was high in both cohorts [93% (78-98) vs 97% (94-99), p=0.34], the rENE+ group had inferior 5- year DC [78% (59-88) vs 95% (91-97), p<0.001)] and DFS [58% (43-77) vs 90% (86-94), p<0.001]. MVA identified rENE+ as the strongest independent adverse prognostic factor for DFS [HR 4.3 (95% CI 2.3-8.1), p<0.001]. T2 (vs T1) category [HR 2.1 (1.0-4.2), p=0.039], smoking pack- years (continuous) [HR 1.02 (1.0-1.03), p=0.013], and addition of systemic agents [HR 0.4 (0.2-0.8), p=0.005] were also prognostic. RPLN, LN number (continuous), and lower neck LN were significant in UVA but not MVA. Conclusion Data from this contemporary TNM-8 cT1-T2N1 HPV+ OPC cohort suggests that the presence of rENE is associated with reduced DC and DFS, whereas the number of radiographically evident LNs is not an independent predictor for DFS. While these results are derived in a high- volume head and neck cancer centre, external validation in a more general setting is desirable before consideration of rENE inclusion in the refinement of future editions of the HPV+ OPC TNM stage classification.