Applying the technique of volume‐modulated arc radiotherapy to upper esophageal carcinoma

This study aims to evaluate the possibility of using the technique of volume‐modulated arc therapy (VMAT) to combine the advantages of simplified intensity ‐ modulated radiation therapy (sIMRT) with that of regular intensity‐modulated radiation therapy (IMRT) in upper esophageal cancer. Ten patients with upper esophageal carcinoma were randomly chosen in this retrospective study. sIMRT, IMRT, and VMAT plans were generated to deliver 60 Gy in 30 fractions to the planning target volume (PTV). For each patient, with the same clinical requirements (target dose prescription, and dose/dose‐volume constraints to organs at risk (OARs)), three plans were designed for sIMRT (five equispaced coplanar beams), IMRT (seven equispaced coplanar beams), and VMAT (two complete arcs). Comparisons were performed for dosimetric parameters of PTV and of OARs (lungs, spinal cord PRV, heart and normal tissue (NT)). All the plans were delivered to a phantom to evaluate the treatment time. The Wilcoxon matched‐pairs, signed‐rank test was used for intragroup comparison. For all patients, compared to sIMRT plans, VMAT plans statistically provide: a) significant improvement in HI and CI for PTV; b) significant decrease in delivery time, lung V20, MLD, heart V30 and spinal cord PRV D1cc; c) significant increase in NT V5; and d) no significant reduction in lung V5, V10, and heart MD. For all patients, compared to IMRT plans, VMAT plans statistically provide: a) significant improvement in CI for PTV; b) significant decrease in delivery time, lung V20, MLD, NT and spinal cord PRV D1cc; c) significant increase in NT V5; and d) no significant reduction in HI for PTV, lung V5, V10, heart V30 and heart MD. For patients with upper esophageal carcinoma, using VMAT significantly reduces the delivery time and the dose to the lungs compared with IMRT, and consequently saves as much treatment time as sIMRT. Considering those significant advantages, compared to sIMRT and IMRT, VMAT is the first choice of radiotherapy techniques for upper esophageal carcinoma. PACS number: 87.55. D‐

in the upper esophageal region, including cervical and upper thoracic esophagus, radiotherapy isanefficienttreatmentselection.Howeverradiotherapytreatmentplanningischallengedby the complex anatomical features of upper esophageal region. The delivery of high radiation dose used in conformal radiotherapy technique is often limited by the tolerance of organs at risk(OARs).Fixed-field,intensity-modulatedradiationtherapy(IMRT)achievesanevenmore conformal dose distribution of target structures, while OARs are spared to a greater extent.
High-modulated IMRT plans using multiple beam angles and complex intensity modulation notonlyresultinaprolongedtreatmenttime,butalsoreducetheefficiencyofradiationbecause of the use of many small control points. (2) Inefficientradiationleadspatientstounnecessary exposure from scatter and leakage dose. Hall and Wuu (3) showed that, compared with the three-dimensional(3D)conformalradiationtherapy,high-modulatedIMRTcausedtheincrease in volume of normal tissue being exposed to a low-dose of radiation, which is estimated to increasetheincidenceofsecondarycancersfrom1%to1.75%attenyears.Thensimplified intensity-modulated radiation therapy (sIMRT) technique was widely used in clinic. (4) sIMRT is introduced to reduce the number of MUs, but the HI and CI of the target are deteriorated, as well as the OAR sparing.
In this case, volumetric-modulated arc therapy (VMAT) was reported as a novel radiation technique, which can achieve highly conformal dose distributions with improved target volume coverage and sparing of normal tissues (NT). Over the past few years, VMAT has been previously compared with IMRT for various types of cancer, (5)(6)(7)(8)(9)(10)(11)(12)(13)(14)(15)(16) and several researches have shown that VMAT has the ability to produce dosimetrically equivalent plans to IMRT for centrally located cancers, such as prostate cancer, cervical cancer, anal canal cancer, and head and neck cancers. (4)(5)(6)(7)(8)(9)(10) In these studies, VMAT was further demonstrated to have the capability of reducing the number of monitor units required to deliver treatments when compared with IMRT, in turn reducing overall treatment times. Martin et al. (17) pointed out that using the arc technique is a good option for mid-and distal-esophageal cancer with primary involvement of the gastroesophageal (GE) junction. However the treatment time hasn't been investigated, and VMAT has seldom been reported in the treatment of upper esophageal carcinoma, which has the complex anatomical features. Therefore, this study aims to evaluate the possibility of using VMAT to integrate the advantages of both sIMRT and IMRT.

B. Plans
For each patient, sIMRT, IMRT, and VMAT plans were designed. The photon beam energy for all plans was 6 MV to be delivered on Elekta Synergy linac (Elekta, Stockholm, Sweden) equippedwith40pairsofleaves.Dosegridresolutionwas0.4× 0.4 × 0.4 cm. For all patients, thebeamarrangementsofsIMRTandIMRTplanswerefiveandsevenequispacednonopposed coplanar beams in 360°, beginning with 0°, respectively. The number of control point was limitedtonomorethan25and45forsIMRTandIMRTplans.Therefore,inthisarticle,the difference between sIMRT and IMRT plans was that there were more two beams and dozens of control points in IMRT plans. The beams would be delivered using a multileaf collimator (MLC) with the step-and-shoot technique.VMAT plans were generated with two coplanar completearcs(dualarc).Deliverytimewasnotlimited.Continuousgantrymotion,dose-rate variation,andMLCmotionwereapproximatedbyoptimizingindividualbeamsat4°gantry angle increments. The same dose-volume constraints were used for all the plans during inverse planning with direct machine parameter optimization method. (19) DeliverysettingsforsIMRT(IMRT)were:minimummonitorunitsforeachsegment,ten (five);minimumsegmentarea,10(5)cm 2 ;andmaximumnumberofsegments,25(45).For VMAT, the gantry angle spacing was 4 0 .Thefinaldosedistributionswerecalculatedwiththe adaptive convolution method. The planning goal was to deliver a prescribed dose of 60 Gy to atleast95%ofPTVin30fractions;thedoseuniformityrequirementwas-5%to+7%.For OARs, V20 for bilateral lungs was no more than 30%, heart V30 was no more than 30%, and themaximumdosedeliveredtothespinalcordPRVwasnomorethan45Gy.
For each pair of sIMRT and VMAT plans, the optimization objectives were the same and originated from IMRT plan. The ultimate goal of optimization for individual patients was that dose to OARs can be kept as low as possible while maintaining optimal target coverage and dose uniformity to the target.
CI (20) isdefinedasfollows: where V T is the target volume, V T,ref is the target volume covered by the reference isodose line, and V ref is the total volume covered by the reference isodose line. The reference dose was 60 Gy. The value of CI is between zero and one. CI = 1 represents the ideal situation that the target volume coincides exactly with the treatment volume; CI = 0 represents a plan in which there is no overlap between the two volumes.
HIisdefinedasthedifferencebetweenthedosescovering5%and95%ofthePTV. (21) The equation is as follows: (2) A greater value of HI indicates a greater degree of dose heterogeneity in the PTV.
The dosimetric parameters of OARs were chosen according to OAR characteristics. For bilateral lungs, heart, and NT, the parameters were the mean dose and the percentage volume thatwasirradiatedatspecificdose(e.g.,V5,V10,andV20forthelungs).Forspinalcord PRV,theparameterD1ccwasthemaximumdosedeliveredto1cm 3 . Plans were delivered to a phantom for time comparison.

A. A representative patient
The dose-volume histogram for planning target volume and OARs of the three different treatment techniques are shown in Fig. 1. The curves show that, when dealing with the low-dose region(e.g.,V5)tothenormaltissue,VMATisworsethansIMRTandIMRT;whendealing with the relatively high-dose region (e.g., V20), VMAT is superior to sIMRT and IMRT; and when dealing with the high-dose region (e.g., V64.2), VMAT performs more excellently than sIMRT and IMRT. Figure 2 shows the isodose distributions in the central axial, and sagittal and coronal planes for one representative patient. It is obvious that: a) for PTV coverage, PTV homogeneity, and lung V20, the VMAT and IMRT plans are similar and better than the sIMRT plan; b) for spinal cord PRV, the high isodose lines can form bigger C-shape region to sparing the spinal cord PRV in the central axial planes in VMAT; c) for normal tissue, the 20 Gy isodose line shows significantlessvolumeintheregionneartospinalcordPRVinthecentralaxialplanesinVMAT; andd)fornormaltissue,the5Gyisodoselineshowssignificantmorevolume.

B. All patients
The results of the study are shown in Table 1 and Table 2. The third column in the table lists thep-values.Ap-valueof<0.05isconsideredclinicallysignificant.
As shown in Table 1, compared to sIMRT plans, VMAT plans statistically provide: a) significant improvement in HI and CI for PTV; b) significant decrease in delivery time, lungV20,MLD,heartV30andspinalcordPRVD1cc;c)significantincreaseinNTV5;and d)nosignificantreductioninlungV5,V10,andheartMD.Therefore,theseresultsindicatethat VMAT keeps the advantage of sIMRT and even achieves better, while sparing the OARs and without losing of the HI and CI, but the volume of low-dose regions for NT is increased.
Compared to IMRT plans (Table 2 provides details), VMAT plans statistically provide: a)significantimprovementinCIforPTV;b)significantdecreaseindeliverytime,lungV20, MLD, and spinal cord PRV D1cc; c) significant increase in NT V5; and d) no significant reductioninHIforPTV,heartV30andheartMD.Consequently,thesecomparisonsshowthat VMAT keeps the advantage of IMRT and even achieves better, but the volume of low-dose regions for NT is increased.

IV. dIScuSSIon
In this study, we compared the VMAT with IMRT and sIMRT. The outcome shows that VMAT has the advantage of saving time, even in comparison with sIMRT, which obviously has the advantage of reducing treatment time. With faster treatment, VMAT can improve the machine efficiency,whilereducingthediscomfortofpatientsandthepossibilityofintrafractionmovements during treatment. (6,14) Besides, theoretically, the prolongation of treatment time has negative implication, although the exact clinical effect is still uncertain. (22) Hence VMAT can avoid these minor issues.
For patients with the tumor situated in the paraspinal region, who frequently due to the spinal cord PRV can't suffer the high dose, coverage of the PTV will be worse and even the prescribed dose will be decreased. However, VMAT indicates stronger ability to control the distribution of high-dose area through intensity modulation and gantry continuous rotation. (6) Therefore, the oncologists can scale up the delivered dose to tumor and provide a potential improvement in treatment outcome. (23) In this work, we also investigated the planning time (data not shown). Compared with IMRT, thetimeperiterationoptimizationandperfinaldosecomputationforVMATisthreeand20 And it's reported that the increased planning time limits the VMAT clinical application. (6,8,12) However, Quan et al. (24) developed an AutoPlan system to automatically generate the VMAT plans with less planning time. In future, more advanced optimization system may emerge to reduce treatment planning time. (25) The results also indicate that VMAT plans perform better in dosimetric parameter indices (V20 and mean lung dose) than sIMRT and IMRT using the same dose constraints. Although theV5andV10inVMATplanswerenotsignificantlyhigherthanthoseinIMRTplans,obviously they got increased when compared to conformal plans, (26,27) which may lead to increase the risk of radiation pneumonitis. Hall and Wuu (3) concluded that IMRT may approximately double the risk of secondary cancers as a consequence of the changing CRT to IMRT because ofabiggervolumeofnormaltissuebeingexposedtolowerdoses.Ourfindingagreeswith the conclusion of Quan et al. (24) that VMAT creates a larger volume of low-dose irradiation to normal tissue than IMRT. In order to learn more about the feature of the low-dose region in VMATplans,dose-volumeconstraintsV5wereassignedtothelungs.Theresultsshowthe volume of low-dose regions could be limited using VMAT. In addition, the analysis shows that the HI of VMAT plans is superior to IMRT plans, which has never been reported before, and this may be attributable to the numerous beam angles, especially in the region of heterogeneous electrondensity.Furtherinvestigationisrequiredtoconfirmtheinference.