6th ICHNO Abstract Book

6th ICHNO 6 th ICHNO Conference International Conference on innovative approaches in Head and Neck Oncology 16 – 18 March 2017 Barcelona, Spain __________________________________________________________________________________________ page 53

lymphoscintigrams were scored as non-SLNs when SPECT- CT was added. There were 4 false negative SLN biopsy procedures in this cohort.

mm, respectively. The distances between all four POs to CTV1 and the 95% isodose curve were consistent. 41% of POs were within CTV1 (Figure 1.) and 58% were covered by the 95% isodose curve. In 13 cases, LRRs were outside the treatment volume. LRR identified 5 mm outside the CTV1 accounted for 13.3% of all LRR and may represent marginal recurrences in the present study. Conclusion We confirmed that the majority of loco-regional recurrences appear in the high dose volume. This is consistent using four different recurrence identification approaches. The current data may differ between centres using larger treatment margins and this is currently under investigation. PO-111 Dermal backflow: NIRFLI pattern associated multimodality therapy in patients with oropharynx cancer S. Naqvi 1 , I. Tan 2 , J. Rasmussen 2 , M. Aldrich 2 , J. Morrow 2 , A. Blanco 3 , C Gutierrez 4 , K. Jain 1 , E. Sevick-Muraca 2 , R. Karni 1 1 The University of Texas Health Science Center- Houston, Otorhinolaryngology- Head and Neck Surgery, Houston, USA 2 The University of Texas Health Science Center- Houston, Center for Molecular Imaging, Houston, USA 3 The University of Texas Health Science Center- Houston, Radiation Oncology, Houston, USA 4 The University of Texas Health Science Center- Houston, Physical Medicine and Rehabilitation, Houston, USA Purpose or Objective Lymphedema is a ubiquitous side effect of oropharynx cancer treatment and is characterized by swelling and fibrosis which leads to a deterioration in quality of life. Currently there is limited understanding about how to properly diagnose and treat head and neck lymphedema. Our team has previously described a novel imaging system for the identification of aberrant lymph vessel flow of the cervical region in patients with Head and Neck Cancer. Material and Methods Following intraoral and intradermal injections of indocyanine green, lymphatic anatomy was visualized using near-infrared fluorescence lymphatic imaging (NIRFLI). Six oropharynx cancer patients were imaged prior to treatment, and then followed up for thirty weeks. One subject received neither neck surgery nor radiation, one patient had bilateral therapeutic-range neck radiation without neck dissection, two patients had unilateral neck dissection with post-operative radiation to the operated neck alone, one patient had unilateral neck dissection but post-operative radiation applied to both sides of the neck, and one patient had bilateral neck dissections and bilateral post-operative neck irradiation. Results Lymphatic vessels were readily visualized in all six subjects (12 necks). All five necks that received both neck dissection and post-operative therapeutic-dose radiation (>56Gy) exhibited dermal backflow (5/5). In the necks that received neither surgery nor radiation, dermal backflow was not observed (4/4). One of three necks that received radiation alone without neck dissection also developed dermal backflow. Conclusion Post-operative radiotherapy is strongly associated with the development of dermal backflow. Likewise, treatment de-escalation in early stage oropharynx cancer which spares a therapeutic dose of radiation to the contralateral neck also preserves the native lymphatic function. It is our contention that NIRFLI is an emerging imaging modality for the identification of aberrant lymphatic backflow following treatment of oropharynx cancer and specifically

Conclusion The addition of SPECT-CT to planar lymphoscintigraphy is recommended for the identification of more (positive) SLNs and better topographical orientation for surgery in sentinel lymph node biopsy for early stage oral cancer. PO-110 Analysis of loco-regional failures after IMRT for HNSCC using deformable image registration R. Zukauskaite 1 , C.R. Hansen 1 , J. Johansen 1 , C. Grau 2 , J. Overgaard 3 , C. Brink 1 1 Odense University Hospital, Department of Oncology, Odense, Denmark 2 Aarhus University Hospital, Department of Oncology, Aarhus, Denmark 3 Aarhus University Hospital, Department of Experimental Clinical Oncology, Aarhus, Denmark Purpose or Objective Most failures after curative radiotherapy for head and neck squamous cell carcinoma (HNSCC) appear loco- regionally. Moreover, IMRT may increase the risk of marginal failures due to the high conformity. Identification and analysis of possible points of loco- regional recurrence (LRR) in relation to treatment volume may pave the way for improved treatment planning and outcomes. The aim of this project was to analyze loco- regional recurrence pattern after curative IMRT. Material and Methods 457 patients with larynx, pharynx and oral cavity HNSCC were consecutively treated with primary IMRT from 2006- 2012 in one institution. The majority of patients had loco- regionally advanced disease. Most patients were treated with curatively intended primary IMRT with 66-68 Gy in 6 fractions weekly with concomitant Nimorazol and Cisplatin according to DAHANCA guidelines. Elective therapeutic levels were treated to 50 Gy. All CT or PET/CT-confirmed loco-regional recurrences were identified, and the diagnostic CTs were deformable registered with planning CTs. Four different approaches to define possible points of LRR origin (PO) were identified: two by independent observers, mass mid-point (MMP), and a point with maximal surface distance (MSD). Points were further analyzed in relation to clinical high-risk treatment volume (CTV1) and planned treatment volume (95% isodose of high-risk volume). Results Median follow-up was 41 months (range 2-107). 290 patients (63%) remained recurrence-free and 167 developed some sort of failure. The three-year loco- regional control rate was 68%. 70 patients with 105 CT verified LRR were further analysed. 35, 19 and 16 pts had recurrences in T, N or both T and N sites, respectively. Median distances between the four different ways to estimate POs in the x, y and z axis were 3.0, 3.0 and 8.2