ESTRO 36 Abstract Book

S890 ESTRO 36 2017 _______________________________________________________________________________________________

The DVCs were then used as limits such that the dose that could be delivered would result in the tightest constraint being just met. Therefore, the dose for that fraction was increased or decreased to ensure that the DVC was on the tolerance limit. The impact of the dose escalation was then evaluated using TCP and NTCP. Results Thirteen of the patients investigated could have received net higher doses during their treatment without exceeding their OAR DVCs. In the remaining 18 patients, only 20 fractions out of 257 would allow an increase in dose while staying below the DVC limits. The rectum was the limiting structure in 97 % of fractions. The largest individual increase possible for a given fraction was 87.4 cGy. If all changes were made, the maximum accumulated net increase in dose possible for any patient was 13.58 Gy, assuming the imaged fractions were representative of the patients’ entire treatment and scaling to a full treatment. This corresponded to an increase in TCP and rectal NTCP of 13.7 % and 13.6 % respectively. Table 1 shows the results for the 13 patients. Conclusion Adapting the dose to be delivered to the patient on a fraction-by-fraction basis has the potential to allow for significant dose escalation while staying within institutional DVCs, significantly increasing TCP. This could be particularly useful in the hypofractionation approach to treatments. [1] Physica Medica, 32(4):618–624, 2016. EP-1661 Adaptive strategy to accommodate anatomical changes during RT in oesophageal cancer patients T. Nyeng 1 , M. Nordsmark 2 , L. Hoffmann 1 1 Aarhus University Hospital, Medical Physics, Aarhus C, Denmark 2 Aarhus University Hospital, Department of Oncology, Aarhus C, Denmark Purpose or Objective During chemoradiotherapy (chemoRT) in oesophageal cancer (EC), some patients show large interfractional anatomical changes. These changes may affect the dose distribution adversely, demanding adaptation of the treatment plan. The aim of this study was to investigate a decision support system for treatment adaptation based on daily cone-beam CT (CBCT) scans. Material and Methods Twenty consecutive patients treated with chemoRT for oesophageal and gastro-oesophageal junction cancer were setup to the spinal cord with a tolerance of 5mm using daily CBCT scans. On CBCT, mediastinal structures are barely visible. Therefore, a surrogate structure (SS) was used to evaluate the actual target position. The SS was generated by indicating the borders between dense tissue nearby the clinical target volume (CTV) and lung tissue or air, see Fig1. Geometrical changes above 3mm in the tissue defined by the SS were registered by the radiation therapists (RTTs) for each fraction. Additionally, the RTTs noted changes of the base line diaphragm position above 5mm, the mediastinum above 5mm, the body contour above 10mm, and the shoulder blades above 10mm. Three consecutive registrations in any category triggered an adaptation of the treatment plan, requiring a new CT-scan with IV contrast. Targets and organs at risk were re- delineated, based on deformably propagated contours from the planning CT-scan. We recalculated the original treatment plan on the new CT-scan to evaluate the consequences of the observed anatomical changes.

Results Thirteen patients had their RT plan adapted at least once during treatment. In the first adaptation (13 pts), the median decrease in the CTV receiving 95% of prescribed dose (V95%) and the planning target volume (PTV) was 0.2% [0-6.4%] and 5.2% [0-11.7%], respectively, see Fig2. The largest underdosage was related to interfractional baseline shifts in the diaphragm position (6 pts), with a median decrease in CTV V95% of 0.6% [0-6.4%] and PTV V95% of 8.6% [1.7-11.7%]. Target deviations registered as changes in SS (8 pts), were typically caused by swelling of the target area (7 pts), shrinkage or swelling of the mediastinum (4 pts) and/or compression or stretching of the target due to changed diaphragm position (1 pt). Changes in SS were pooled and showed a median decrease in CTV V95% of 0.2% [0-0.6%] and PTV V95% of 5.0% [0.5- 6.9%].