Abstract Book

S276

ESTRO 37

Material and Methods

reduction was feasible and has led to smaller irradiated volumes. OC-0523 Individual dose limitation should be given for left ventricular myocardium in thoracic radiotherapy Y. Tong 1,2 , G. Gong 1 , Y. Yin 1 J.Lu 1 1 Shandong Cancer Hospital affiliated to Shandong University, Radiation Physics Department, Jinan, China 2 University of South China, Nuclear Science and Technology, Hengyang, China Purpose or Objective The dosimetric parameters evaluation for the heart and sub-structures held too much uncertainties which were caused by respiration and cardiac activity. In this study, we achieved the quantification of variation rules in dose- volume parameters for left ventricular myocardium caused by cardiac activity during thoracic tumor radiotherapy. Material and Methods The left ventricular myocardium of 15 patients were contoured on 20 phases of electrocardiography (ECG)- gated four-dimensional computed tomography (4DCT) images in breath-hold. Radiotherapy plans were designed on planning 3DCT images, dose distributions of the plans were imported into MIM Maestro (MIM) and were deformed to each phases. Then the dose-volume parameters of left ventricular myocardium among 20 phases were obtained. Variations in volume and dice similarity coefficient (DSC) for left ventricular myocardium were analyzed. And then the variation in D mean , V 10 , V 20 , V 30 and V 40 for left ventricular myocardium among 20 phases were compared. Results (1) The maximum rate of variation in volume and

Twenty-seven consecutively treated rectal cancer patients were evaluated. Patients received a 5x5Gy radiation dose, on-line position verification and a post treatment Cone Beam CT scan (CBCT). Patients were instructed to have a full bladder and an empty rectum. The radiation oncologist delineated the GTV and CTV including: mesorectum, presacral area and pelvic lymph nodes. A library of plans was created with 5 CTVs, see figure A, and numbered 1 (the most empty rectum) till 5 (the most full rectum) using in-house developed software with an algorithm based on population statistics. Fourteen RTTs were trained to select a plan based on the varying daily anatomy, see figure B,C and D. In total 270 plans were selected, before and after treatment. The plan selections were evaluated on plan selection difference between pre- and post CBCTscans, overall number of selected plans per fraction, and the limiting factor in selecting a smaller plan. Out of 135 treatment fractions, the selected plan on pre and post CBCT was different in 12 treatment fractions. This was due to the change in rectal- (7) and bladder (5) fillings; both smaller (5) and larger (7) plans than on the pre CBCT were selected. The overall number of plans per fraction was dominated by plan 3, which is equal to planning CT (pCT). Plan 3 was selected 95 times, plans 2,4,5 in 8, 22, 10 times, respectively (see table). Plan 1 was removed from the library after the first 10 patients, since it was never chosen, mainly due to its very small volume. We also evaluated the limiting factor in selecting a smaller plan. The size and position of the GTV was limiting in 168 cases. Forty five times the rectal filling was limiting, of which 24 times due to air, 19 times due to feces, and 10 times a combination of both. The bladder was limiting in 5 cases. Two post CBCT scans were missed due to great urgency to empty the bladder. For the first 27 patients, all CBCT scans had a sufficient quality to select a plan. In none of the treatments a back-up plan (plan 4, equal in PTV volume to treatment method without LoP) was used. The RTTs managed the plan selection very well, they only consulted the physician twice. Conclusion Evaluation of the first 27 patients showed a practical library of plans strategy for rectal cancer patients that was easily executable by RTTs. The 1cm PTV margin Results

DSC for the left ventricular myocardium reached to 24% and 184%, and the difference of the maximum and minimum value were statistical significant (p<0.05). (2) The maximum rate of variation in Dmean for the left ventricular myocardium was 87%, the biggest value reached 164%, and the difference of them had statistical significance (p<0.05). (3) The maximum value of variation in V10, V20, V30 and V40 were 14%, 13%, 12% and 4% for the left ventricular myocardium during cardiac cycle, and the difference of them had statistical significance (p<0.05)(Figure 1).

(2)

(3) Conclusion

The variation in volume, morphology and dose-volume parameters for the left ventricular myocardium during cardiac cycle were too significant to be ignored, which suggested that the individual dosimetric limitation must be given for left ventricular myocardium in thoracic tumor radiotherapy.