21 How can we improve treatment plan quality by statistical reporting of patient data?. (December 2018)

• Record Type: Journal Article
• Authors: Robilliard, M.
• Is Part Of: Physica medica. Volume 56(2018)Supplement 1
• Page Start: 13
• Publication Date: 2018-12
• Subjects: Medical physics -- Periodicals; Biophysics -- Periodicals; Biophysics -- Periodicals; Imagerie médicale -- Périodiques; Radiothérapie -- Périodiques; Rayons X -- Sécurité -- Mesures -- Périodiques; Physique -- Périodiques; Médecine -- Périodiques; 610.153
• DOI: 10.1016/j.ejmp.2018.09.034
• Languages: English
• ISSNs: 1120-1797

Abstract

Abstract : Introduction: A statistical reporting is intended to guarantee the results of treatment plans and to help detect the change of practice. The aim of this study is to obtain an indicator in order to know if the optimization was successful and to maintain a good clinical practice. Methods: In our department, the medical prescription for patients treated for prostate cancer is a 75 Gy dose (2.5 Gy per fraction) on the prostate and a 46 Gy dose (2 Gy per fraction) on the iliac lymph nodes. Plans with VMAT (arcs of 6 MV, dose rate: 600 UM/min) are calculated in the Eclipse Treatment Planning System (AAA algorithm, Varian) version 13.6. The coverage of the 95% of PTV with 97% of the prescribed dose is always respected. With the analysis of 32 treatment plans, DVH values for organs at risk (OAR) are collected and are treated in the ascending order. -Bladder: V57Gy, V70Gy, V75Gy (%) -Bladder wall: V67Gy (cc) -Rectum: V46Gy, V60Gy, V70Gy, V75Gy (%) -Rectum wall: V67Gy (cc) These values are divided into four quartiles: Q1, Q2, Q3, and Q4. For each plan, a homemade dosimetric score is computed. If dosimetric indice is smaller than Qn, the score is n. If dosimetric indice is between n and (n + 1), the score is (n + 1). An average score on each OAR is calculated: 1 is the best and 4 the worst achievable score. These scores have been determined for 45 treatment plans calculated with Progressive Resolution Optimizer (PRO13) algorithm and 45 treatment plans calculated with Photon Optimizer (PO13) algorithm. A paired Student's t-test ( α = 0.05) was used to cross-check all plans. Finally, the possibility of improving a plan was studied for a plan calculated with PO13 and a score higher than 4. Results: Results reported in Table 1 show our values of Q1, Q2, Q3, and Q4. With the new algorithm PO13, the average score is reduced from 3.8 to 3.2, with a more important benefit on the rectum because the score is decreased in average from 4 to 3.2. And, the differences were significant (p < 0.05). Preliminary results indicate the possibility of reducing a score higher than 4, by increasing constraints on the rectum during the optimization (dose and priority). With the first treatment plan, the rectum score was able to be decreased from 4.6 to 3.2. Conclusions: Combined with the dosimetric indices, these scores are useful for improving our clinical practice. Our service has set new standards: now, values Q2 are our optimization constraints and we continue to optimize if the score is higher than 4. The values of quartiles will be reviewed if you change your practice, with an improvement. The second aim of this study is to configure a knowledge-based planning (Rapidplan) with a robust and representative sample of plans.