Supervisors info:
Καλλιόπη Πλατώνη, Επίκουρη Καθηγήτρια, Ιατρική Σχολή, ΕΚΠΑ, Επιβλέπουσα
Βασίλειος Κουλουλίας, Αναπληρωτής Καθηγητής, Ιατρική Σχολή, ΕΚΠΑ
Ευστάθιος Ευσταθόπουλος, Καθηγητής, Ιατρική Σχολή, ΕΚΠΑ
Summary:
Purpose: To calculate the tolerance limits and action limits for intensity-modulated radiation therapy (IMRT) and volumetric-modulated arc therapy (VMAT) patient-specific quality assurance (QA), according to the recommendations of TG-218, to examine the clinical relevance of gamma passing rates (%GP) and to compare the two-dimensional (2D) %GP with the three-dimensional (3D) ones, according to the dose errors (%DE) detected by them.
Materials and Methods: Patient-specific QA was performed for one hundred head and neck (H&N) treatment plans and seventy-three prostate cancer treatment plans, delivered with IMRT and VMAT techniques. The measurements were acquired with the cylindrical phantom, ArcCHECK, and the points of the dose distributions where compared using the γ index method. For the tolerance and action limits calculation, thirty prostate plans and thirty H&N plans, randomly chosen from the clinic’s database, were analyzed. The measurements were collected sequentially on the same day and the two-dimensional %GP were computed with SNC Patient software, using the gamma criteria of 3%/3mm and 3%/2mm. The field size and the gradient measure where, then, correlated with the %GP of all the plans. The dose-volume histogram (DVH) of fifty-three prostate cancer treatment plans and seventy-five H&N cancer treatment plans were reconstructed with the 3DVH software, to detect the %DE between the compared dose distributions, while at the same time, the three-dimensional %GP were collected for each structure. The 2D %GP as well as the 3D %GP were correlated with the %DE and, finally, the comparison of the two kinds of %GP was performed according to the true positive (TP), false negative (FN), false positive (FP) and true positive (TP) cases that were observed with each one of them, using a 90% action limit for the %GP and a 3% action limit for the %DE.
Results: The tolerance limits and the action limits calculated for the 3%/2mm gamma criterion were in concordance with the ones recommended in TG-218, regarding the prostate cancer cases, but they were lower for the H&N cases. The degradation of the %GP is partially explained by the small field of the cases that did not pass the QA test and the small value of their corresponding gradient measure. Both the 2D %GP and the 3D %GP where highly correlated with the %DE of DVH values for PTV structures using the 3%/2mm gamma criterion, but not with the %DE for the organs at risk (OAR). By comparing the two kinds of %GP, more TP cases were observed with the 3D ones but, also, more FN cases, than the ones observed with the 2D %GP.
Conclusions: The deviation of the tolerance limits and action limits that were calculated from the ones recommended, reveals the need to improve the accuracy of our IMRT and VMAT treatment plan calculations. The lack of high intensity correlation of the %GP with the dose deviations of the DVH values for the OAR, confirms the fact that gamma passing rates should not be used independently for patient-specific IMRT QA, but in combination with the clinically relevant DVH parameters.
Keywords:
Action limits, Tolerance limits, Patient-specific QA, IMRT, Gamma passing rate, ArcCHECK, 3DVH
