Abstract Book

S313

ESTRO 37

4 university hospital saarland, department for nuclear medicine, homburg saar, germany 5 helios klinikum wuppertal, department for nuclear medicine, wuppertal, germany 6 marienhospital stuttgart, department for radiation oncology, stuttgart, germany 7 marienhospital stuttgart, department for nuclear medicine, stuttgart, germany 8 mutterhaus der boromaeerinnen, department for radiation oncology, trier, germany 9 university hospital magdeburg, department for radiation oncology, magdeburg, germany 10 charite university hospital, department for radiation oncology, berlin, germany 11 university hospital mainz, department for radiation oncology, mainz, germany 12 gemeinschaftspraxis fuer strahlentherapie, department for radiation oncology, neumuenster, germany 13 helios kliniken schwerin, department for nuclear medicine, schwerin, germany 14 university of mainz, institute of medical biometry epidemiology and informatics, mainz, germany Purpose or Objective To evaluate the impact of 18-FDG-PET/CT based target volume delineation on outcome in radiochemotherapy for locally advanced (LA) non-small cell lung cancer (NSCLC). Material and Methods A prospective randomized phase II international multicenter non-inferiority study was performed including 205 of 311 screened patients with LA NSCLC scheduled for combined radio-chemotherapy in 23 trial sites. Target volumes were 1:1 randomized between a conventional (adjuvant mediastinum to 50 Gy plus dose escalation volume of tumor, atelectasis, FDG- and CT- positive nodal stations; arm A) and an experimental (escalation volume FDG-positive tumor and nodal stations only; arm B) approach. In every patient, isotoxic dose escalation was performed respecting normal tissue constraints (pre-defined in analogy to RTOG 0617) to achieve the individually highest achievable doses between 60 and 74 Gy/2 Gy. Stringent pro- and retrospective imaging- and RT-QA was done during recruitment and follow-up. Simultaneous chemotherapy was given, mainly (88%) using platinum and vinorelbine. Primary endpoint was loco-regional progression, main secondary endpoints were in- and outfield- progression- free-, overall survival and toxicity. After a minimum/mean follow up of 6/22 months, we here report the first results. Results In the well balanced patient group of 205 patients, 172 were treated as per protocol, being mainly in UICC-stage IIIa (36%) and IIIb (56%) with GTVs of in mean 105 ml. Dose escalation (realized by 3-D-CRT (50%) or IMRT (50%)) reached in mean 65.3 Gy (A) and 67.3 Gy (B; p=0.007). Loco-regional progression (LRP) was more frequent in arm A (conventional target volume) with a cumulative incidence of 0.29 / 0.39 vs. 0.14 / 0.20 (arm B) after 12 / 24 months (p=0.078). In stratified Cox-models, independent prognostic factors for LRP were study center (p=0.046) and tumor volume (GTV) (p=0.022) while study arm did not reach statistical significance. The 2-year overall survival was 57% (A) vs. 54% (B; n.s.). Neither for local control nor for survival, a positive or negative influence of radiotherapy dose was observed. The first site of tumor progression were distant metastases (A: 56% vs. B 72%; n.s.) followed by the primary tumor (A: 44% vs. B: 26%, p=0.032), in-field- (A: 23% vs. B: 11%; n.s.) and out-field- (A: 11% vs. B: 8%; n.s.) regional nodes. Acute and late toxicity (CTC/ RTOG-EORTC) was mild with no clinically significant difference between the treatment arms. However, SUEs appeared more frequent in arm B.

Purpose or Objective Tumor-infiltrating lymphocytes (TILs) represent a surrogate marker for the adaptive immune system. A minor subset of breast cancers (10-11%) has been found to have a high number of TILs, and increasing evidence has shown that high level of TILs is associated with improved recurrence free- and overall survival (OS) in especially triple negative- and HER2 positive breast cancers. High level of TILs has further been found to predict response to chemotherapy in breast cancer. Our aim was to investigate the predictive value of TILs in terms of benefit from radiation therapy (RT) in pretreatment tumor material from a cohort of breast cancer patients randomized to RT or not. Material and Methods The Danish Breast Cancer Group (DBCG) 82bc cohort constitutes high risk breast cancer patients (tumor size > 5 cm and/or positive lymph nodes and/or invasion in skin or pectoral fascia) diagnosed between 1983-89, treated with mastectomy and axillary lymph node dissection followed by adjuvant systemic treatment and randomized to +/- post-mastectomy radiotherapy (PMRT), and associated with > 20 years of clinical follow-up. In 1,011 pretreatment, tumor-containing paraffin blocks from the DBCG82bc cohort, stromal TILs were estimated by 2 observers using hematoxylin-eosin staining’s following international recommendations. A competing risk model, Kaplan-Meier analysis and multivariate Cox regression analysis (MVA) were used for analyzing correlations between TILs and clinical outcome. Results Patients were dichotomized into groups of 'low” and 'high” levels of TILs using a 30% cut off, leading to 10.5% of patients (106/1,011) being categorized as having 'high” level of TILs. 'High” TILs level was identified in MVA as an independent prognostic factor for lower risk of distant metastasis (DM) (adj. HR= 0.70, CI: 0.53-0.94) and improved OS (adj. HR=0.73, CI: 0.57-0.94), but was not found to be prognostic for loco-regional recurrence (LRR). In both TILs groups a benefit from PMRT could be found for all clinical endpoints. The benefit of PMRT to reduce the risk of LRR was found to be equal in the two TILs groups at 20 years (24% actuarial risk reduction for 'low” TILs vs. 26% for 'high” TILs, test for interaction: p=0.45). A trend for greater reduction was observed for DM (11% vs. 19%, p=0.29). A significantly greater benefit from PMRT was observed for 'high” TILs patients as compared to 'low” TILs patients with significantly improved OS at 20 years (8% improvement for 'low” TILs (from 23% to 31%) vs. 22% for 'high” TILs (from 26% to 48%), p=0.045). The association remained significant when adjusting for estrogenreceptor- and HER2 status. Conclusion The results showed that high levels of pretreatment TILs interacted with RT to further improve OS, but had no impact on local control in a cohort of breast cancers. The finding may indicate that local RT through tumor destruction and release of tumor antigens triggers a local immune response that induces a systemic effect outside the treatment field (abscopal effect). OC-0598 FDG-PET-guided target volume reduction for isotoxic dose escalation in LA NSCLC (PET-Plan study) U. Nestle 1 , T. Schimek-Jasch 2 , S. Kremp 3 , A. Schaefer- Schuler 4 , A. Kuesters 1 , M. Tosch 5 , T. Hehr 6 , S. Eschmann 7 , Y. Bultel 8 , P. Hass 9 , T. Brunner 9 , J. Fleckenstein 3 , A. Thieme 10 , M. Stockinger 11 , M. Miederer 11 , M. Beck 12 , G. Holl 13 , H. Rischke 2 , S. Adebahr 2 , S. Lenz 14 , J. Koenig 14 , A. Grosu 2 1 Kliniken maria hilf gmbh, department of radiation oncology, mönchengladbach, germany 2 university hospital freiburg, department for radiation oncology, freiburg, germany 3 university hospital saarland, department for radiation oncology, homburg saar, germany