ESTRO 37 Abstract book

ESTRO 37

S358

Conclusion In routine clinical practice GTV CT and GTV MR demonstrated only limited concordance. The addition of MRI co-registration into the clinical pathway lead to a GTV MRCT which was significantly larger. MRI and CT potentially provide complimentary information for accurate GTV delineation, and this may lead to a lower risk of marginal treatment failures. However, with the absence of a pathological ‘gold-standard’ for these patients, correlation of imaging with recurrence patterns is essential to avoid unnecessary increases in target volumes. PO-0703 Added value of metastatic cervical lymph node group V in nodal staging of nasopharyngeal cancer P. Chaysiri 1 , J. Setakornnukul 1 , K. Thephamongkhol 1 1 Siriraj Hospital- Mahidol University, Radiation Oncology, Bangkok, Thailand Purpose or Objective We hypothesized cervical lymph node group V to be the 2 nd echelon group, same as lymph node group IV, due to its possibility of receiving lymphatic drainage from anterior cervical group. This study aims to determine the added prognostic significance of cervical lymph node group V beyond the standard AJCC staging system of nasopharyngeal cancer patients treated with IMRT in terms of overall survival. Material and Methods Retrospective cohort of 199 consecutive diagnosed nasopharyngeal cancer patients treated with definitive radiotherapy or concurrent chemoradiotherapy (CCRT) using intensity-modulated radiotherapy (IMRT) technique in single university hospital in endemic area. Pre- treatment imaging were newly re-evaluated by using radiographic criteria for cervical lymph node metastasis. Restaging was performed for both T and N staging according to latest 8 th AJCC staging system. Group V cervical node was divided into upper (Va)and lower group (Vb) and evaluated for its prognostic significance according to the 8 th AJCC system. Cox’s proportional hazard model was used to retrieve Hazard ratio (HR), 95% confidence interval and p value. Results For N classification, comparing survival of N3 disease to N0 by using subgroup of lower cervical group V(Vb) only or combination of cervical group IV and Vb (supraclavicular fossa replacement according to 8th AJCC system) as N3, our study showed greater separation of survival by using group Vb only with HR of 6.75 (95%CI 1.94-23.51, p=0.003) compared to HR of 4.70 (95% CI 1.37-16.13, p=0.014) by using current standard N3 disease (group IV and Vb). Interestingly, the overall survival (OS) in N2 disease with group Va is nearly similar to N3 disease without cervical group Vb, and in the same way, N1 disease with group Va to N2 disease without group Va (Table 1). After incorporating cervical group V into nodal staging, the HR and 95% CI among each group was better separated than the 8 th AJCC staging system (Table 2 and Figure 1).

(CT) scan has the potential to improve the accuracy of GTV delineation for primary tumours. Deformable coregistration of a diagnostic MRI to the planning CT has been implemented clinically in our centre in 2015. This study aims to evaluate the impact of this change in clinical practice upon primary tumour GTV delineation. Material and Methods 23 consecutive patients with histologically proven HNSCCs (21 oropharynx, 1 nasopharynx and 1 hypopharynx) treated with radical (chemo)radiotherapy were identified following the clinical implementation of deformable coregistration of a T1w MRI in 2015 in our centre, in whom co-registration had been deemed clinically acceptable by visual assessment and all primary tumour GTV delineations had been carried out as part of routine practice by a single radiation oncologist. MRI images were deformably registered to geometrically align with CT. All contours were performed viewing all available diagnostic imaging (including MRI) side-by-side. For each patient as part of routine clinical practice three GTV contours had been manually delineated: GTV on the CT only without coregistration (GTV CT ), GTV on the T1-MRI only (GTV MR ), and a final GTV based upon both the GTV CT and GTV MR (GTV CTMR ). Differences between contours were quantitatively analysed using volumetric and geometric comparison metrics. The five comparison metrics used were volume difference (%), centre of gravity distance (CGD), DICE similarity co-efficient, mean distance to conformity (MDC) and conformity index (CI). Wilcoxon signed rank test was used to test for statistically significant differences between the comparison metrics produced. Results Distribution of contour volumes are displayed in Figure 1. GTV CT and GTV MR were both significantly smaller than the GTV CTMR (P<0.01). There was no significant difference in volume between GTV CT and GTV MR (P=0.69). Contour comparison metrics are summarised in Table 1. GTV CT and GTV MR showed limited overlap (DICE =0.71). Overlap indices demonstrated greater concordance between both GTV CT and GTV MR with GTV CTMR (DICE = 0.84 and 0.82 respectively) than to one another. Neither GTV CT nor GTV MR were individually significantly closer to the GTV CTMR than the other for any of the contour comparison metrics measured.