P1.14-017 Impact of Systematic EBUS-TBNA Mediastinal Staging on Radical Radiotherapy Planning in NSCLC

S2042

Journal of Thoracic Oncology

Vol. 12 No. 11S2

7.64, respectively. Mean FLT-PET SUVmax for groups 1 and 2 was significantly different (p¼0.03) at 3.42(1.14-7.04) and 2.34(1.234.35) respectively. Similarly, mean (range) of SUV50, SUV95 and SUVmean for group-1 was 1.8(0.74-3.43), 2.97(1.03-5.83), 1.87(0.733.44), and for group-2 was 1.22(0.81-2.26), 1.85(1.13-3.8) and 1.25(0.83-2.39), respectively (p<0.01, <0.01 and <0.01). There was no statistically-significant difference between SUV2Dpeak and SUV3Dpeak between groups 1 and 2, with a mean of 2.97(0.99-6.30) and 2.91(0.90-6.11) for group-1 and 2.10(1.11-3.91) and 2.03(1.003.86) for group-2 (p¼0.06 and 0.06), respectively. There was no statistically significant difference between the 3D and 4D image acquisition in group-1. There were no FLT-PET-related toxicities. Conclusion: FLT-PET is feasible in SBRT patients pre- and posttreatment, and may assist in distinguishing fibrosis from recurrent tumor. Further validation studies are needed. Keywords: Early stage non-small cell lung cancer, radiotherapy, stereotactic

P1.14-017 Impact of Systematic EBUS-TBNA Mediastinal Staging on Radical Radiotherapy Planning in NSCLC

P1.14-016 Assessing the Feasibility of FLT-PET for Evaluation of Non-Small Cell Lung Cancer (NSCLC) Treated with Stereotactic Body Radiotherapy (SBRT) H. Raziee, A. Hope, A. Bezjak, A. Sun, J. Cho, J. Bissonnette, D. Vines, B. Driscoll, M. Giuliani Radiation Oncology, University of Toronto and Princess Margaret Cancer Center, Toronto, ON/CA Background: Distinguishing fibrosis from tumor recurrence following lung SBRT remains a clinical challenge since CT has poor sensitivity and specificity for detecting recurrence. 18F-Fluoro-Lthymidine-PET (FLT-PET) uptake correlates with cell proliferation. The purpose of this study is to investigate the feasibility of FLTPET as an imaging biomarker for lung SBRT response assessment. Method: In this prospective study, three groups were included: 1) newly-diagnosed biopsy-proven NSCLC pre-SBRT, 2) established post-SBRT mass-like fibrosis on serial follow-up CT scans by coinvestigators’ consensus, and 3) biopsy-proven locally-recurrent NSCLC after SBRT. Non-gated, helical gated (3D-CT/4D-PET) and phase-matched (4D-CT/4D-PET) FLT-PET images were obtained. Group-1 underwent fluorodeoxyglucose (FDG)-PET scan according to clinical guidelines. FLT uptake was measured by SUV95 and SUV50 (95% and 50% of maximum pixel value plus average background value, respectively), SUV2Dpeak and SUV3Dpeak (1cm diameter circular or spherical around region of interest, respectively), SUVmean and SUVmax. Descriptive statistics were gathered. KolmogoroveSmirnov test was used to determine normality. Statistical significance was reported using student’s t-test. Result: 27 patients were included, with 19 primary tumors (group-1), 12 established fibrosis (group-2) and 1 recurrence (group-3). In group-1, 16 tumors were T1. Group-1, mean FDG-PET SUVmax, SUV95, SUV50, SUV2Dpeak, SUV3Dpeak and SUVmean were 7.40, 5.88, 2.39, 5.59, 6.02 and 2.78, respectively. Mean FLT-PET values for group-1 were 3.43, 2.84, 1.71, 2.9, 2.82 and 1.78, respectively. Group-2 SBRT dose was either 48Gy in 4 fractions (83%) or 60Gy in 8 fractions. Median time from radiation to FLT-PET scan in group-2 was 19.5 months (5.883.8mos). The patient in group-3 had SUV50, SUV95, SUV2Dpeak, SUV3Dpeak, SUVmean and SUVmax of 2.27, 3.85, 6.37, 6.05, 2.39 and

N. Hardcastle,1 A. Cole,2 G. Turgeon,2 R. Thomas,2 B. Gearey,2 L. Irving,3 B. Jennings,4 T. Kron,1 D. Ball,2 D. Steinfort,3 S. Siva2 1 Physical Sciences, Peter MacCallum Cancer Center, Melbourne, VIC/AU, 2 Radiation Oncology, Peter MacCallum Cancer Center, Melbourne, VIC/ AU, 3Respiratory Medicine, Royal Melbourne Hospital, Melbourne/AU, 4 Respiratory Medicine, Monash Medical Center, Clayton/AU Background: Radical radiotherapy often relies solely on radiological imaging to determine treatment volumes. Systematic mediastinal staging with endobronchial ultrasound transbronchial needle aspiration (EBUS-TBNA) may identify PET-occult sites of mediastinal disease, or demonstrate benign causes for PET-positive LN. This study evaluated 1) Involved nodal coverage 2) Doses to organs-atrisk when planned based on PET-CT and EBUS-TBNA and 3) Incident dose to mediastinal nodes between 3D-CRT and IntensityModulated-Radiotherapy (IMRT). Method: Radical radiotherapy plans (60Gy/30 fractions) were created for patients with stage change following EBUS-TBNA from a prospective clinical trial. We compared lung Normal-Tissue-Complication-Probability (NTCP, pneumonitis), oesophageal and heart dose for planning to targets based on PET-CT versus PET-CT+EBUS-TBNA. The incidental dose to PET-negative/EBUS-TBNA-positive nodes from 3DCRT and IMRT was evaluated using volume receiving 35Gy as a surrogate for control of sub-clinical disease (Kepka, IJROBP, 73(5) 2009). Result: Of 30 patients enrolled, four were upstaged by EBUS-TBNA; these patients had a significant geographic miss of nodal GTV when planned to PET-positive nodes only (Figure 1). When planned based on PET-CT alone, the incidental dose to PET-negative/EBUSTBNA-positive nodes was higher with IMRT for two patients (v35Gy increased by 17% & 6%; Figure 1a&b) and lower with IMRT (v35Gy reduced by 16% and 6%; Figure 1c&d) for two, dependent on nodal position relative to the primary. Six patients had negative pathology for PET avid nodal stations; Inclusion of EBUS-negative, PET-positive nodes resulted in an average increased lung NTCP of 5% (range 1%-13%), mean oesophagus dose of 13Gy (range 4-23Gy) and mean heart dose of 4Gy (range -0.1-11Gy) over plans based on EBUS-positive nodes alone. Conclusion: Systematic EBUS-TBNA has the potential to improve loco-regional control and limit the probability of lung and heart toxicity. The incidental dose to adjacent tissue is inherently related to involved node/tumor position and not solely dictated by the radiation delivery technique. Keywords: EBUS-TBNA, NSCLC, TCP/NTCP

November 2017

Abstracts

S2043

model and that 100 plans is more effective. When compared with manually optimized plans the M_COM model demonstrates an improved relationship between balancing OAR objectives and PTV conformity. Keywords: Radiotherapy, VMAT, Improvement

P1.15-001 Ipilimumab Increases Th1/Th2 and Inflammatory Cytokines Counteracting Chemotherapy Effects in Small Cell Lung Cancer M. Hardy-Werbin,1 P. Rocha,2 O. Arpí,1 Á. Taus,2 D. Joseph-Pietras,3 A. Rovira,1 J. Albanell,2 C. Ottensmeier,3 E. Arriola2 1Cancer Research Programm, Hospital Del Mar Medical Research Institute, Barcelona/ES, 2 Medical Oncology Department, Hospital Del Mar, Barcelona/ES, 3Nihr Experimental Cancer Medicine Center, University of Southampton, Southampton/GB

P1.14-018 Construction and Evaluation of RapidPlanTM Models for Radical Lung Planning Y. Flanagan,1 P. Kelso,1 S. Smith,1 T. Mitchell2 1Physics, Beatson West of Scotland Cancer Center, Glasgow/GB, 2Beatson West of Scotland Cancer Center, Yn/GB Background: A Volume Modulated Arc Therapy (VMAT) service for our radical lung patients has been offered since 2011. Attention has now turned towards how this service can be improved further. The RapidPlanTM package has already been used to great effect in our Prostate service so was assessed for its potential in radical lung planning. Method: RapidPlanTM models were created using the Eclipse Treatment Planning System [Varian Medical Systems] v13.6: consisting of 50 randomly selected radical right lung patients (M_R), left lung patients (M_L) and a combined model using all 100 of these same patients (M_COM). The models were assessed using the standard RapidPlanTM tools. A retrospective planning study was performed for 10 right lung and 10 left lung patients. Each of the 20 patient treatments was replanned using each model. All plans were normalized to ensure that the target mean was 100% and compared to the manually optimized plan (O_P) The PTV coverage was assessed using dose homogeneity indices. Differences in organ at risk (OAR) dose volume histogram parameters were calculated to assess plan quality. The plans were compared on a variety of criteria including but not limited to whole lung V5 (with tolerance considered 70%) and V20 (with tolerance considered 30%) and the maximum dose to spinal canal (tolerance considered to be 80%). Significance was assessed by two-tailed ttest (p<0.01). Result: Although all models gave a clinically acceptable plan differences in the homogeneity indices were observed with M_COM values showing an improvement on the other models. M_COM plans had a statistically significant increase in maximum spinal cord dose when compared to the other plans set, however doses were within tolerance. An increase in mean whole lung and whole lung minus PTV V20Gy was also observed. There was no significant difference in contra lateral lung V5Gy or whole lung minus PTV V5Gy. There was no statistical difference observed when comparing right sided tumor patient calculated with M_L and left sided tumor patients planned with M_R. Conclusion: M_COM outperformed the other models and O_P in relation to PTV coverage without impacting clinically acceptable OAR constraints. This model demonstrated improvement when compared to M_L and M_R suggesting that 50 plans are insufficient to perfect a lung

Background: Cytokines are soluble proteins with a relevant role in immune response that can be modulated by immunotherapy. In this study we aimed to evaluate the serum concentrations of Th1/Th2 and inflammatory cytokines and their changes in SCLC patients treated with ipilimumab and chemotherapy compared to chemotherapy alone. Method: We evaluated 2 cohorts (C) of patients with SCLC: in C1, 47 patients were treated with standard platinum/etoposide (PE); in C2, 37 patients with ipilimumab, carboplatin and etoposide (ICE trial). We analyzed serum samples at baseline and subsequent time points with the Cytokine Human Magnetic 10-Plex (GM-CSF, IFN-g, IL-1b, IL-2, IL-4, IL-5, IL-6, IL-8, IL-10 and TNF-a) plus MIP-1a. Serum samples from 30 healthy volunteers were used as controls. Statistical analysis was carried out with SPSS 22.0 (SPSS Inc.) and Prism 6.0 (GraphPad). Result: Patients with SCLC showed significantly lower levels of IL-1b, Mip-1a, IL-5 and IL-10 and higher TNF-a compared to healthy volunteers. Patients treated with chemotherapy alone (C1), showed a decrease of Th1 and inflammatory cytokine median concentrations at tumor response and an increase at progression, while no such pattern was observed in Th2 cytokines and TNF-a. In contrast, patients treated with ipilimumab in addition to chemotherapy (C2), showed a global increase on the level of all cytokines after treatment initiation (Fig 1). Conclusion: In patients with SCLC, Th1 and inflammatory cytokines (except TNF-a) reflect tumor burden. Chemotherapy reduces these levels targeting cytokine secreting tumor cells. Th2 cytokines may be mainly secreted by immunological cells as they remain unaltered upon chemotherapy treatment. Interestingly, ipilimumab increases globally Th1, Th2 and inflammatory cytokine secretion counteracting the effect of chemotherapy. The correlation of these changes to outcome and toxicity warrants further investigation. Keywords: small cell lung cancer, cytokines, Ipilimumab