ESTRO 38 Abstract book

S551 ESTRO 38

tumoral- and normal appearing brain tissue after combined immunotherapy and SRS. Material and Methods Twenty-eight patients from our on-going prospective observational imaging study ( clinicaltrials.gov Identifier: NCT03458455 ), with a total of 37 brain metastases from non-small cell lung cancer (N=14, n=17) or malignant melanoma (N=14, n=20), have so far been included. All patients received linear accelerator-based SRS (15-27Gy). The dose was prescribed to cover at least 99% of the PTV; contrast enhancing tumor on T1w post-contrast MRI+2 mm isotropic margin. The imaging protocol included pre-SRS MRIs and follow-up MRIs every third month for one year. Six patients (5 melanomas, 1 lung) received immunotherapy before and after SRS, and five patients (2 melanomas, 3 lung) started immunotherapy within 6 months post-SRS. Immunotherapy, pembrolizumab (2mg/kg) or ipilimumab (3mg/kg), or a combination, was given every third week. Relative cerebral vessel calibers (rVSI), and relative blood volume (rCBV) from micro- (spin echo) and macrovasculature (gradient echo), were calculated from perfusion MRI and Vessel Architectural Imaging. All values were normalized to normal appearing white matter receiving less than 2Gy. Normal brain tissue was defined as white- and gray-matter segmented from T1w pre-contrast images in SPM12 (Matlab), excluding tumor and edema which were drawn by experienced neuroradiologists. The normal appearing brain tissue was divided into five ROIs according to doses: >0-2Gy, 2>-5Gy, 5>-10Gy, 10>-15Gy and >15Gy (peri-tumoral region) (Fig.1).

were compared, as well as PTV coverage. All patients’ treatment plans were delivered without the use of functional information. Results Majority of patients had stage III disease (67%). Preliminary analysis of first eight patients showed the mean PTV volume of 375 cc, anatomical lung volume of 3595 cc. The mean volume for FL40 was 1144 cc. The ratio of functional lung to anatomical lung volume was 0.32. There was weak correlation between anatomical lung and FL volumes within patients (Pearson r=0.2), due to perfusion defects variations (tumour, emphysema). The largest dose reduction was achieved with IMRT plans (Figure 1). Plans optimised to FL subvolumes resulted in a significant dose reduction by a mean of 2.1 Gy (20.2%) to the highest functional subvolume FL80 (CI 0.4-3.9 p=0.02). Dose reduction to FL40 and FL60 was 1.6 Gy (CI 0.6-2.7) and 1.5 Gy respectively (CI 0.3-2.8 p=0.02). Functional V20 improved by 5.4% (CI 1-10 p=0.03). Max dose to OAR (heart, oesophagus and spinal cord) and PTV coverage were not significantly different between plans. Detailed analysis of the whole patient cohort is underway.

Conclusion Functional avoidance planning optimised to perfused lung volumes identified with SPECT resulted in improved dose volumetric outcomes for functional lung. This methodology may lead to potential reduction in radiation- induced lung toxicity in patients treated with definitive lung RT, and subsequently offers potential for target dose escalation. PO-1000 Vascular responses in normal brain tissue after combined immunotherapy and SRS to brain metastases L. Nilsen 1 , E. Grøvik 2 , I. Digernes 2 , C. Saxhaug 3 , A. Latysheva 4 , O. Geier 2 , T.P. Hellebust 5 , D.O. Sætre 6 , B. Breivik 7 , K.D. Jacobsen 8 , Å. Helland 8 , K.E. Emblem 2 1 Oslo University Hospital, Diagnostic Physics, Asker, Norway ; 2 Oslo University Hospital, Diagnostic Physics, Oslo, Norway ; 3 Oslo University Hospital, Radiology and Nuclear Medicine, Olso, Norway ; 4 Oslo University Hospital, Radiology and Nuclear Medicine, Oslo, Norway ; 5 Oslo University Hospital, Medical Physics, Oslo, Norway ; 6 Østfold Hospital Trust, Radiology, Kalnes, Norway ; 7 Hospital of Southern Norway, Radiology, Kristiansand, Norway ; 8 Oslo University Hospital, Oncology, Oslo, Norway Purpose or Objective Stereotactic radiosurgery (SRS) is a well-established treatment option for patients with brain metastases. Over the past few years, there has been a rapid increase in the additional use of immunotherapy. At present, there is limited understanding of the mechanisms behind the combined effects of immunotherapy and SRS. Our aim is to provide deeper insight into vascular responses of peri-

Results Before SRS, there were no significant differences in rVSI or micro- and macrovascular rCBV between the two treatment groups in any of the dose-level ROIs (Fig.2). However, six months post-SRS, patients treated with SRS and immunotherapy showed higher relative increase in rVSI (1.4;0.91-2.1 vs 0.82;0.51-0.97 (p<0.01)) together with greater reduction in micro-vascular rCBV (0.67;0.45- 0.93 vs 1.1;0.98-1.3(p<0.05)) in the peri-tumoral region. Median macro-vascular rCBV tended to decrease in both treatment groups, and there was no difference in the relative change (0.92;0.66-1.2 vs 0.85;0.68-0.96 (p=0.5)).