ESTRO 36 Abstract Book

S241 ESTRO 36 2017 _______________________________________________________________________________________________

Purpose or Objective Advances in intracranial stereotactic radiotherapy have led to high gradient dose between tumor and normal tissue and to dramatically reduced Planning Target Volume (PTV) margins. Accurate definition of the gross tumor volume (GTV) for stereotactic radiotherapy of brain metastases is an essential key for the treatment planning. However, its underestimation due to tumor growth during the delay between planning and stereotactic radiotherapy may lead to treatment failure. Our purpose was to evaluate the tumor growth kinetics and its impact during the delay before treatment of brain metastasis secondary to lung cancer (LC) or melanoma (ML). Material and Methods This retrospective monocentric study included all consecutive patients (pts) treated for brain metastases secondary to LC or ML between June 2015 and May 2016. Margins from GTV to PTV were 2 mm. Imaging at diagnosis of brain metastasis and preplanning imaging were compared; GTV corresponding to the contrast enhancement was analyzed. Linear extrapolation was used to determinate the n minimum theoretical time leading the diameter of the tumor to increase more than 4 mm (T4mm). Results Out of 103 pts treated for brain metastasis by stereotactic radiotherapy, 50 were treated for metastases secondary to LC (n=26) or ML (n=24). Six pts were excluded because of lack of imaging data. Median age was 68 years old (range: 25-92). RPA status was 1 for 1 patient (2%), 2 for 33 pts (79%) and 3 for 8 pts (19%). Systemic treatment was given at diagnosis for 19 pts (45 %). Radiotherapy was delivered according to a monofraction scheme for 8 pts (3 LC and 5 ML metastasis), 3-fraction scheme (23 LC, 18 ML) or 5-fraction scheme (2 LC, 3 ML). A hundred and eight brain imaging (84 MRI, 24 CT-scan) were analyzed. Comparison of imaging at diagnosis and preplanning treatment showed bleeding inside metastasis for one patient with primary LC; increased tumor volume for 40 pts (ML n=25 ; LC n=15) ; stability for 11 pts (ML n=1 ; LC n=10) and decreased volume for one LC patient. Median delay between brain imaging at diagnosis and pretreatment planning were: 28 days (range 8-107) for ML pts and 31.5 days (range 7-70) for LC pts. Median Volumes of GTV at diagnosis were 0.5 cm3 (range 0.05-8.6cm 3 ) for ML pts and 0.45cm 3 (range 0.05-6.1cm 3 ) for LC pts; median volumes of preplanning treatment GTV were 1.55 cm 3 (range: 0.2-9.9cm 3 ) for ML pts and 0.85 (range 0.2- 10.4cm 3 ) for LC patients. Linear extrapolation revealed a median increase of tumor volume of 0.16 cm3/wk (range 0-0.8 cm3/wk) for ML and 0.06 cm3/wk (range 0-0.5 cm3/wk) for LC. Shorter T4mm was 15 days for ML patients and 17 days for LC pts. Conclusion Maximal delay for treatment appeared to be 15 days for ML patients and 17 days for LC patients to ensure that tumor radius has grown less than to 2 mm. Above this delay, clinicians should reconsider planning of treatment. PV-0458 FMECA of Cyberknife process: two years’ experience for improvement S. Cucchiaro 1 , D. Dechambre 1 , T. Massoz 1 , N. Gourmet 1 , D. Boga 1 , N. Jansen 1 , P. Coucke 1 , M. Delgaudine 2 1 C.H.U. - Sart Tilman, Radiotherapy Departement, Liège, Belgium 2 C.H.U. - Sart Tilman, STA Quality Departement, Liège, Belgium Purpose or Objective Failure Modes Effects and Criticality Analysis (FMECA) is a risk analysis allowing the identification of causes and effects of a potential problem and the prioritization of actions that can reduce this dysfunction. Our Radiation Therapy Department used the FMECA as a strategy tool to

continuously improve treatment quality and safety. This FMECA approach was applied to our Cyberknife (CK) workflow process. Material and Methods Using the FMECA methodology, the CK workflow process was defined with a flow chart and responsibility map including a description of every step of prescription, treatment preparation and treatment delivery. The identification of possible risks was then carried out with their origins and consequences. The evaluation was based on 3 criteria: Severity (S), frequency of Occurrence (O) and probability of Detection (D). Finally, we calculated the Criticality Index (CI = S x O x D) for each of the identified risks. The rating for each criterion is based on a scale from 1 to 4. The Criticality Index can span a range of 1 to 64. Results We defined 10 stages, with corresponding failure modes presented in a table. At each stage, identified failures with possible causes and consequences are listed and the risk level assessed. A detailed scoreboard was obtained presenting the risks and enabling easier identification of priority actions to be undertaken. The board showed 66 possible failure modes. 8 of the top-ranked failure modes were considered for process improvements. We also crossed the scoreboard obtained with the adverse events most often reported on 2015. We found 2 correspondences between failure modes and adverse events reported. We therefore also considered that in the implementation of preventive/improvement actions to take. A review of this analysis was done in September 2016. Therefore, at this moment, a revaluation of the process, failures, ratings and implemented actions was performed with each members of the CK team. The correlation with reported adverse events was also made. We had one failure mode that has to be changed from a moderate to an unacceptable level because an incident was reported following a non-update procedure. New improvement actions have been implemented directly. In order to continue our proactive approach to risk analysis a systematic annual review of this analysis is now introduced in routine. All this, in relation to the reported adverse events. The figure shows an extract of the FMECA scoreboard obtained for CT simulation and contouring stage.