Abstract Book

ESTRO 37

S363

O’Sullivan 1 , J. Ringash 1 , A. Hansen 3 , J. De Almeida 4 , E. Monteiro 4 , D. Chepeha 4 , R. Gilbert 4 , J. Irish 4 , D. Goldstein 4 , J. Waldron 1 , A. Hope 1 1 Princess Margaret Cancer Centre, Department of Radiation Oncology, Toronto, Canada 2 Princess Margaret Cancer Centre, Department of Biostatistics, Toronto, Canada 3 Princess Margaret Cancer Centre, Department of Medical Oncology, Toronto, Canada 4 Princess Margaret Cancer Centre, Department of Otolaryngology-Head & Neck Surgery/Surgical Oncology, Toronto, Canada Purpose or Objective To investigate the predictors and outcomes for oral cavity squamous cell carcinoma (OSCC) patients (pts) treated with surgery who had early recurrence (ER) before planned postoperative radiotherapy (PORT). Material and Methods Retrospective review was conducted for all OSCC pts treated with PORT between 2003-2015 following curative- intent gross total resection. Pts were classified according to the timing of recurrence and intent of PORT; those who had no recurrence before PORT were treated with adjuvant intent (adjuvant-group), while pts who had ER before PORT were treated with salvage intent if no distant metastases (DM, salvage-group) or palliative intent if DM were detected before initiating PORT (palliative-group). ER was detected clinically by CT and/or MRI, and pathologic conformation was used when imaging finding was indeterminate. For the whole cohort, multivariable analysis (MVA) using logistic regression was applied to identify predictors of ER. In the salvage group, the success of PORT to control the ER diseases (i.e. salvage rate) was estimated by calculating the proportion of pts who were disease free following salvage PORT, and univariable analysis (UVA) was used to identify factors associated with locoregional control (using competing risk regression), disease free- (DFS) and overall survival (OS) using cox proportional hazard regression (tested variables included: interval between surgery and PORT, volume of ER gross disease, type of treatment failure before PORT, and use of concurrent chemotherapy with PORT) Results Overall 601 pts were identified; of whom 513 (85%) were treated with adjuvant PORT, while 88 (15%) had ER (28/88, 32% were biopsy-proven) before PORT (70 in the salvage-group and 18 in the palliative-group). Tumor and treatment characteristics are summarized in table-1. The 3-year OS for OSCC pts treated with adjuvant, salvage and palliative PORT were 71% (95%CI: 67%-75%), 41% (95%CI: 30%-56%) and 0% (95%CI: 0%-0%) respectively (p<0.001). On MVA, oral tongue subsite (HR=1.8, 95%CI:1.1-3.1, p=0.02), microscopic positive resection margin (HR=2.4, 95%CI:1.4-4.1, p=0.003), pT3-4 category (HR=2.9, 95%CI:1.7-4.9, p<0.001) and pN2-3 category (HR=5.7, 95%CI:2.9-11.2, p<0.001) were associated with development of ER. Among the salvage group, the overall salvage rate was 37% (26 pts out of 70, remained disease free with a median follow up of 4.8 years), and on UVA larger gross tumor volume was associated with poor local control, DFS and OS (p=0.007, 0.02, and 0.04 respectively).

Conclusion Baseline postoperative imaging is recommended to detect ER in OSCC pts, especially those with pT3-4, pN2-3, positive microscopic margin, and oral tongue subsite. Pnts with ER have poor survival, however substantial proportion of them could be properly salvaged with PORT+/-concurrent chemotherapy. Future studies are required to improve prediction of ER, and investigate experimental therapy to improve the outcomes. PO-0710 Brain dose from IMRT is associated with neurocognitive function in nasopharyngeal cancer survivors L. McDowell 1 , J. Ringash 1 , B. Chan 1 , W. Xu 2 , L. Lu 2 , J. Waldron 1 , K. Rock 1 , S.H. Huang 1 , N. So 1 , M. Giuliani 1 , A. Hope 1 , B. O'Sullivan 1 , S. Bratman 1 , J. Cho 1 , J. Kim 1 , R. Jang 3 , A. Bayley 1 , L. Bernstein 4 1 Princess Margaret Cancer Centre, Department of Radiation Oncology, Toronto, Canada 2 Princess Margaret Cancer Centre, Department of Biostatistics, Toronto, Canada 3 Princess Margaret Cancer Centre, Division of Medical Oncology, Toronto, Canada 4 Princess Margaret Cancer Centre, Department of Supportive Care, Toronto, Canada Purpose or Objective To determine whether dose from incidental brain irradiation is prognostic for cognitive or neurobehavioural dysfunction in long-term nasopharyngeal carcinoma (NPC) survivors treated with intensity-modulated radiotherapy (IMRT). Material and Methods Eligible NPC patients were (1) disease-free; (2) ≥ 4y after IMRT completion; and (3) ≥18y at time of irradiation. Objective cognitive function was assessed with the Montreal Cognitive Assessment (MoCA) using a cut-off of 23 (<23 v ≥23) to define cognitive impairment. Patient- reported cognitive function was assessed using the MDASI- HN “problems with remembering” item. Frontal lobe behaviours (apathy, disinhibition, executive dysfunction) were assessed with the total and behavioural subscale scores from the Frontal Systems Behaviour Scale (FrSBe). Temporal lobe necrosis (TLN) was recorded on follow-up MRI imaging. Brain structures were contoured using CT and MRI imaging and absolute volumes were reported at 5Gy increments from V5 to V75 for: (1) left temporal lobe (LT); (2) right temporal lobe (RT) and (3) both frontal lobes (FL). Results FL-V60 (OR=0.01, 95% CI: 0-1.14, p=0.057) and V55 (OR=0.27, 95% CI: 0.06-1.13, p=0.073) showed a trend in predicting cognitive impairment (MoCA score <23 vs. ≥23) on univariate logistic regression. As a continuous variable, total MoCA showed weak correlation with RT- V25 (R=-0.18, p=0.07) and V20 (R=-0.18, p=0.075). Visuospatial scores correlated weakly (-0.20>R>-0.26) but significantly (p<0.05) with RT-V10 to V20 and LT-V15 to V35. Abstraction scores showed weak correlation with RT- V75 (R=-0.24, p=0.02). Orientation scores were predicted