SU‐E‐T‐24: A Simple Correction‐Based Method for Independent Monitor Unit (MU) Verification in Monte Carlo (MC) Lung SBRT Plans

Abstract

Purpose:Lung‐SBRT uses hypo‐fractionated dose in small non‐IMRT fields with tissue‐heterogeneity corrected plans. An independent MU verification is mandatory for safe and effective delivery of the treatment plan. This report compares planned MU obtained from iPlan‐XVM‐Calgorithm against spreadsheet‐based hand‐calculation using most commonly used simple TMR‐based method.Methods:Treatment plans of 15 patients who underwent for MC‐based lung‐SBRT to 50Gy in 5 fractions for PTV V100%=95% were studied. ITV was delineated on MIP images based on 4D‐CT scans. PTVs(ITV+5mm margins) ranged from 10.1‐ 106.5cc(average=48.6cc). MC‐SBRT plans were generated using a combination of non‐coplanar conformal arcs/beams using iPlan XVM‐Calgorithm (BrainLAB iPlan ver.4.1.2) for Novalis‐TX consisting of micro‐MLCs and 6MV‐SRS (1000MU/min) beam. These plans were re‐computed using heterogeneity‐corrected Pencil‐Beam (PB‐hete) algorithm without changing any beam parameters, such as MLCs/MUs. Dose‐ratio: PB‐hete/MC gave beam‐by‐beam inhomogeneity‐correction‐factors (ICFs):Individual Correction. For independent‐2nd‐check, MC‐MUs were verified using TMR‐based hand‐calculation and obtained an average ICF:Average Correction, whereas TMR‐based hand‐calculation systematically underestimated MC‐MUs by ∼5%. Also, first 10 MC‐plans were verified with an ion‐chamber measurement using homogenous phantom.Results:For both beams/arcs, mean PB‐hete dose was systematically overestimated by 5.5±2.6% and mean hand‐calculated MU systematic underestimated by 5.5±2.5% compared to XVMC. With individual correction, mean hand‐calculated MUs matched with XVMC by – 0.3±1.4%/0.4±1.4 for beams/arcs, respectively. After average 5% correction, hand‐calculated MUs matched with XVMC by 0.5±2.5%/0.6±2.0% for beams/arcs, respectively. Smaller dependence on tumor volume(TV)/field size(FS) was also observed. Ion‐chamber measurement was within ±3.0%.Conclusion:PB‐hete overestimates dose to lung tumor relative to XVMC. XVMC‐algorithm is much more‐complex and accurate with tissues‐heterogeneities. Measurement at machine is time consuming and need extra resources; also direct measurement of dose for heterogeneous treatment plans is not clinically practiced, yet. This simple correction‐based method was very helpful for independent‐2nd‐check of MC‐lung‐SBRT plans and routinely used in our clinic. A look‐up table can be generated to include TV/FS dependence in ICFs.