Setup Errors Of Radiation Therapy For Esophageal Carcinoma With On-board KV-CBCT And Its Effects On Dose Distributions

Objective: To study the size of linear setup error during the radiationtherapy for esophageal carcinoma under the guidance of the kilovolt conicalbunch of CT (KV-CBCT) and to analyze the difference of dose of targets andnormal organs in which the first setup error of each radiotherapy.Methods: Using airborne imaging system of Elekta linearaccelerator(XVI),14esophageal carcinoma patients were scanned byKV-CBCT after the initial positioning, after the correction of setup error andafter treatment of every three dimensional conformal radiotherapy(3D-CRT)or intensity modulated radiotherapy(IMRT), the system automaticallyreconstructs sectional image, matched with the form of gray with CT scanningimages of treatment planning to obtain the patients' setup error of threedirections, left-right(use X axis represents), superior-inferior(use Y axisrepresents), anterior-posterior(use Z axis represents). Individual setup errorwas offset distance of three directions in before and after positioning, whichcan be divided into individual systematic error and individual random error.The mean value of all inter-fractional errors expressed the individualsystematic error(X). Standard deviation of all inter-fractional errors expressedthe individual random error(S). Setup errors of all patients expressed thepopulation systematic error, which can be divided into population systematicerror and population random error. Standard deviation of the individualsystematic error expressed the population systematic error (Σ) and standarddeviation of the individual random error expressed the population randomerror (σ). To calculate the margin the formula described by vanHerk was used:PTV margin=2.5Σ+0.7σ.We got the data which was setup error of the first positioning by XVIwhen accepting radiotherapy. According to the first setup error, isocenter ofthe original treatment plan (plan1) moved to the practical position which was the new isocenter under all the remaining same conditions, such as the rayangle, the beam number, beam shape, the prescription dose, multi-leafcollimator, the machine hops, etc., and treatment plan didn't re-optimized.According to the first setup error by CBCT, isocenter of the original treatmentplan (plan1) moved to the practical position which was the new isocenter, gotdose curve of each time the actual radiation and merged each actual radiationdose to obtain the curve of total actual dose at the end of treatment as the newtreatment plan (plan2). Compared plan1and plan2to get the difference ofdose of targets and normal tissues and analyze setup errors of all the firstpositioning to affect on dose distributions when esophageal carcinoma patientsradiotherapy.Result:14esophageal carcinoma patients were scanned787times byCBCT. To calculate the margins through data of setup errors of12esophagealcarcinoma patients by the formula PTV margin=2.5Σ+0.7σ, the conclusionwere as follows: Without the image guide, the margins were X=4.8mm,Y=6.1mm, Z=2.6mm (Table4); with the usage of the image guide, themargins were X=1.7mm, Y=2.8mm, Z=1.7mm (Table5). Applying pairedt-test statistical method to analysis the difference of the margins on X,Y,Zdirections under IGRT and non-IGRT in12esophageal carcinoma patients, theconclusions were as follows: P values follows by0.001,0.001,0.040in X, Y,Z directions (Table7); the margins of X, Y, Z directions were consideredsignificance(P <0.05).Data of parameters which were mean lung dose, V5, V10, V20, V30,V40, mean heart dose, V30, the maximum dose of spinal cord,95%PTVvolume and95%CTV from dose volume histograms of plan1and plan2of14esophageal carcinoma patients (Table8,9,10), applied paired t-test statisticalmethod, the conclusions were as follows: P values follows by0.944,0.055,0.135,0.446,0.005,0.021,0.976,0.571,0.286,0.001,0.009(Table11); MLD,V5, V10, V20, MHD, V30and the maximum dose of spinal cord in before andafter the correction of setup errors were considered non-significance(P>0.05);V30and V40of lung,95%PTV volume and95%CTV were considered significance(P<0.05).Conclusion:1During radiotherapy, the spatial distributions of setup errors werelargest in Y direction, moderate in X direction and smallest in Z direction inwhich all cases of esophageal carcinoma in this group.2Radiotherapy of esophageal carcinoma has a certain degree of setuperror. The correction of setup errors can significantly reduce theinter-fractional setup error, improve the precise of radiotherapy, reduce theCTV-PTV margin, reduce lung injury and avoid the target volume to leak, soeach image guide before radiotherapy is crucial.3In our radiotherapy centers, real-time on-line correction of setup errorsby CBCT to obtain the margins: X=1.7mm, Y=2.8mm, Z=1.7mm, so externalexpansion3mm as three-dimensional direction through image guidance isappropriate.