ESTRO 2023 - Abstract Book

S738

Monday 15 May 2023

ESTRO 2023

R1 in 64 % (n=7/11), and R2 in 9.1% (1/11). Of the remaining, one had a complete response (CR) and proceeded to a wait and watch approach, and four proceeded to palliative treatment due to liver metastasis (n=1) or progression of the local tumour with no surgical offer at re-evaluation (n=3). The highest reported acute haematologic toxicity was grade 2 anaemia (n=2). Post-RT grade ≥ 3 toxicity was reported for three patients (subileus, dysphagia, and urinary tract infection), and post-OP ≥ grade 3 toxicity was reported for two patients (wound infection and pelvic abscess). At a median follow-up of 54 months, the median RFS was 15.8 months (95 % CI, 9.2-23.2). Median OS from the first fraction of re-RT was 47 months, with a 1, 3, and 5-year survival rate of 81.3 %, 56.3 %, and 25.7 %, respectively. The median OS in patients with CR/R0/R1 was 51 months compared to 12 months for R2/non-resected patients. Conclusion Neo-adjuvant hyperfractionated re-irradiation of LRRC is feasible. Due to the manageable acute toxicity dose escalation could be pursued within clinical trials to enhance oncological outcomes. PD-0889 Functional lung dose and postoperative lung complications in esophageal cancer trimodality therapy P. Populaire 1,2 , G. Defraene 3 , P. Nafteux 4,5 , L. Depypere 4,5 , K. Haustermans 1,2 1 Univeristy Hospitals Leuven, Department of Radiotherapy, Leuven, Belgium; 2 KU Leuven, Laboratory of Experimental Radiotherapy, Leuven, Belgium; 3 Laboratory of Experimental Radiotherapy, KU Leuven, Leuven, Belgium; 4 Univeristy Hospitals Leuven, Department of Thoracic Surgery, Leuven, Belgium; 5 KU Leuven, Department of chronic diseases and metabolism (CHROMETA), laboratory for respiratory diseases and thoracic surgery (BREATHE), Leuven, Belgium Purpose or Objective In trimodality therapy for esophageal cancer (EC), pulmonary postoperative complications (pPOCs) are related to the lung dose of neo-adjuvant chemoradiotherapy (nCRT). Meanwhile, studies in lung cancer have shown improved toxicity risk predictions when using dose metrics of the functional lung volume (FLV) instead of the anatomical lung volume. 4D-planning CTs can be used to identify the FLV leading to “functional avoidance planning strategies”. This exploratory analysis examines the role of FLV dose of nCRT for EC as a predictor of pPOCs. Materials and Methods Patients treated in UZ Leuven between 01-2018 and 06-2022 with nCRT for EC using 4D CT-simulation were included in this analysis. Patients were treated to 45Gy in 25 fractions using IMRT, concomitant chemotherapy consisted of carboplatin- paclitaxel. Scans and treatment plans were imported and analysed in MIM Maestro. Total lung volume (TLV) was delineated on the maximum expiration phase. Jacobian determinants based on vector fields of the deformable image registration from maximum inspiration to maximum expiration respiratory phases were calculated (Figure 1 A+B) . FLV was determined based on a 16% threshold of the maximum Jacobian determinant, analogous to the methodology described by Nyeng et al. (Acta Oncol. 2021) (Figure 1 C+D) . Doses of the clinical treatment plan’s average image were non-rigidly registered to the maximum expiration phase (Figure 1 C+D) . Volumes and derived absolute (mL) and relative (%) dosimetric parameters of the TLV and FLV (V40Gy, V30Gy, V20Gy, V10Gy, V5Gy, MLD) were extracted. Parameters were then compared between patients with and without pPOCs, using unpaired t-tests (normal distribution) or Mann-Whitney U-tests (non-normal distribution) using IBM SPSS. Figure 1: Functional lung mapping

Results Out of 51 patients, 12 patients with pPOCs were identified. Statistical analysis showed that the TLV was lower in the pPOCs group when compared to the non-pPOCs group (3141mL (SD: 858mL) vs 3601mL (SD: 635mL) p=0.025 ). The proportion of the TLV identified as functional (ratio FLV/TLV) was significantly higher in patients with pPOCs (29% (SD: 13%) vs 19% (SD: 11%) p=0.024 ). Finally, absolute (mL) FLV 10Gy and FLV 20Gy were significantly higher in patients with pPOCs (FLV 10Gy : 360mL (SD: 216mL) vs 264mL (SD: 139mL) p=0.038 ; FLV 20Gy : 166mL (SD: 106mL) vs 118mL (SD: 63mL) p=0.030 ). There were no statistical differences in relative nor absolute TLV 10Gy (p= 0.326 and p=0.336) and TLV 20Gy (p=0.159 and p=0.444) between both groups. Conclusion Our exploratory analysis shows that lower TLV is associated to the risk of developing pPOCs . Our findings also indicate that the absolute volume of high-functional lung that is exposed to a low to intermediate dose level (FLV 10Gy and FLV 20Gy ) is prognostic of pPOCs . This can lead to improved toxicity prediction models and toxicity-limiting lung avoidance planning strategies for esophageal cancer.