Radiotherapy and Oncology xxx (2018) xxx–xxx

Contents lists available at ScienceDirect

Radiotherapy and Oncology

journal homepage: www.thegreenjournal .com

Original article

Optimal image guided radiation therapy strategy for organs at risksparing in radiotherapy of the prostate including pelvic lymph nodes

https://doi.org/10.1016/j.radonc.2018.02.0090167-8140/� 2018 The Authors. Published by Elsevier B.V.This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

⇑ Corresponding author at: Department of Radiotherapy, Catharina Hospital,Michelangelolaan 2, 5623 EJ Eindhoven, The Netherlands.

E-mail address: aniek.v.nunen@catharinaziekenhuis.nl (A. van Nunen).

Please cite this article in press as: van Nunen A et al. Optimal image guided radiation therapy strategy for organs at risk sparing in radiotherapyprostate including pelvic lymph nodes. Radiother Oncol (2018), https://doi.org/10.1016/j.radonc.2018.02.009

A. van Nunen ⇑, P.P.G. van der Toorn, T.C.G. Budiharto, D. SchuringDepartment of Radiotherapy, Catharina Hospital, Eindhoven, The Netherlands

a r t i c l e i n f o a b s t r a c t

Article history:Received 21 December 2017Received in revised form 9 February 2018Accepted 11 February 2018Available online xxxx

Keywords:Prostate and pelvic lymph nodesOAR sparingImage guided radiotherapyTreatment planningSimultaneous integrated boost

Background and purpose: Purpose of this study was to quantify the OAR dose for different position correc-tion strategies, and to determine which strategy is most optimal for treating patients on the prostate andpelvic lymph nodes.Materials and methods: For 30 patients, four different treatment plans were made reflecting different cor-rection strategies: online correction on bony anatomy; offline correction on bony anatomy; online correc-tion on the prostate fiducials; using 1 cm margins around both CTVs. The dose to the PTVs and OARs wasquantified and a pairwise statistical analysis was performed.Results: No statistically significant differences were observed in the dose to the PTVs, ensuring that anyOAR sparing is not caused by differences in PTV coverage. Dose to the rectum and anal canal was lowestwhen applying an online correction on prostate fiducials, although the total PTV volume was higher. Doseto the small bowel bag and femoral heads was slightly higher compared to online correction on bonystructures, but well within clinically acceptable limits.Conclusion: Although the total PTV volume is higher when applying an online correction on the prostate,this strategy leads to the most optimal sparing of relevant OARs, at the cost of a slightly higher dose to thefemoral heads and small bowel bag.� 2018 The Authors. Published by Elsevier B.V. Radiotherapy and Oncology xxx (2018) xxx–xxx This is anopen access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

The treatment for lymph node positive prostate cancer, oftenconsists of radiotherapy combined with Androgen DeprivationTherapy (ADT). The target volume includes the pelvic lymph nodes(PLN), prostate and seminal vesicles (SV).

Ensuring that the target volume receives the planned dose canbe a challenge due to the fact that the PLN and prostate can movewith respect to each other due to changes in rectal filling [1–3].Different correction strategies can be chosen to ensure eithercorrect positioning of the prostate or the pelvic nodes, resultingin different CTV-PTV margins for both target volumes. Largermargins and treatment volumes results in higher OAR dose andhigher toxicity [4,5]. In a previous study, the required setupmargins for different correction strategies were calculated basedon pre-treatment CBCT images of 20 patients [6]. Calculation wasbased on the Van Herk formula [7,8]. This study showed that whena patient setup correction is performed on a match of the bonyanatomy, larger margins are required for the prostate to accountfor baseline shifts with respect to the PLN. These baseline shiftscan be accounted for by applying an online setup correction on

the prostate, but this will lead to larger PTV margins around thePLN. In a similar study, Kershaw et al. recently quantified theprostate motion in relation to the PLN and SV to define the setupmargins [9]. They concluded that neither a prostate or bone regis-tration is optimal, and that further research is required to deter-mine which strategy results in the lowest OAR doses. A numberof studies have investigated the coverage of the target volumesfor a given PTV margin, but did not report on OAR dose [1,10,11].Hsu et al., Rossi et al. and Eminowicz et al. investigated differentcorrection strategies, but also focused on target coverage. A dosi-metric analysis of these different correction strategies quantifyingthe OAR dose has not been published yet. The purpose of this studywas to determine which setup correction strategy leads to the low-est OAR dose, and consequently, the lowest toxicity.

Materials and methods

Patients

30 consecutive patients with lymph node positive prostatecancer were included retrospectively. All patients received a lymphnode dissection and were eligible for radiotherapy when a maxi-mum of 3 micro-metastases or 2 macro-metastases in the PLNwere found.

of the

2 Radiotherapy for prostate and lymph nodes

Imaging and delineation

All patients received a planning CT scan (Philips Healthcare,Best, The Netherlands) in supine position, with a slice thicknessof 3 mm. Patients were immobilised using a knee and ankle sup-port. To reduce chance of systematic errors due to prostate shiftson the planning CT, patients were instructed to empty their blad-der and defecate one hour before scanning and each treatmentfraction and after that to drink 500 mL of water. If a large rectal fill-ing was observed, the planning CT was repeated after the defeca-tion. The CT scan was imported in the RayStation TPS (version5.0.2.35, Raysearch Laboratories, Stockholm, Sweden). The CTV ofthe prostate (CTVp) consists of the entire prostate gland and theSV were included in this CTV. Delineation of the CTV of the PLN(CTVPLN) was based on consensus recommendations from theRTOG [12]. Bladder, small bowel bag (SBB), femoral heads, rectumand anal canal were delineated. The SBB was defined as the peri-toneal space in which the small bowel can move. The rectumwas contoured from the ischial tuberosities to the rectosigmoidjunction and the anal canal was defined from anus to the musculuspuborectalis. Femoral heads were defined up to and including thetrochanter minor.

CTV-PTV margins

The CTV-PTV margins used in this study are shown in Table 1and depend on the position correction strategy:

A. online correction on bony anatomy;B. offline Shrinking Action Level (SAL) correction protocol on

bony anatomy (n = 3, a = 10) [13];C. online correction on the prostate fiducials;D. CTV-PTV margin of 1 cm around both CTVp and CTVPLN

The setup margins were determined in a previous study, andaccount for translational patient setup errors as well as baselineshifts between the prostate and PLN [6]. Depending on the correc-tion strategy, this results in either larger margins to the prostate orthe PLN to account for these baseline shifts. An additional marginwas included to account for rotation and delineation uncertainties.For a bony anatomy match, both an online and offline correctionwas included as the positional errors in the pelvis are usuallysmall, however an offline strategy will lead to a reduced onlineworkload and a decreased imaging frequency. The CTVs and PTVsfor all evaluated strategies are visualised in Fig. 1.

Treatment planning

For each patient, one treatment plan was made for each of thefour PTV margins reflecting the different correction strategies. Allplans were created for an Elekta linac with an Agility MLC (Elekta,Crawley, UK) using a single VMAT beam with a full gantry rotationand a beam energy of 10 MV. The isocentre was placed in the geo-metric centre of the total PTV volume (PTVp + PTVPLN). The pre-scribed dose was 50.4 Gy to PTVPLN and 67.2 Gy to PTVp in 28fractions using a simultaneous integrated boost (SIB) technique.

Table 1CTV-PTV margins for different correction strategies.

Strategy Prostate margins (cm)

LR SI A

A 0.6 0.9 1B 0.8 1.1 1C 0.5, with 0.8 to apex prostate and around SV [2]D 1.0 1.0 1

Please cite this article in press as: van Nunen A et al. Optimal image guided rprostate including pelvic lymph nodes. Radiother Oncol (2018), https://doi.org

A standard set of objectives was used for optimising the treat-ment plans. The objectives were adapted for the individual patientto achieve an optimal treatment plan. Dose criteria that shouldminimally be achieved for both the PTVs and OARs are summarisedin Table 2 [5]. For the PTVs, 95% of the PTV volumemust be coveredby at least 95% of the prescription dose (V95% �95%). Care wastaken that the PTV coverage was identical for all plans. All gener-ated treatment plans were reviewed by the same expert medicalphysicist and radiation oncologist. Priority was given to reducingthe rectum and anal canal dose, followed by the SBB, as this is clin-ically most relevant. Sparing of the femoral heads and bladder wasconsidered to be less important.

Data collection and analysis

The dose criteria given in Table 2 were evaluated. In addition,the absolute volumes of the different PTVs (PTVp, PTVPLN andPTVtotal), the maximum and the mean dose to the PTVs, rectumand bladder were evaluated. All parameters were automaticallyretrieved using the scripting functionality.

For data analysis, SPSS was used. A Shapiro–Wilk test was per-formed to check if the data were normally distributed. If data werenormally distributed, the mean and standard deviation were calcu-lated. Otherwise the median including minimum and maximumvalue was determined. A pairwise statistical analysis was per-formed on the data to see if a statistically significant differencein OAR dose could be observed between the different strategies,using a repeated measurements ANOVA test or a Friedman test,depending on the normality of the data. When a Friedman testwas used and a significant difference was obtained, a WilcoxonSigned Rank test was applied to point out between which correc-tion strategies the dose was statistically significant different. ABonferroni correction was applied to correct for multiple testing.

Results

Four treatment plans with margins reflecting different correc-tion strategies were optimised. Patient characteristics are sum-marised in Table 3. The mean or median values of the evaluatedparameters obtained from the treatment plans and the results ofthe pairwise statistical analysis are summarised in Table 4.

Differences in the average PTV volumes are statistically signifi-cant between the four strategies (p < 0.001). An online correctionon the prostate results in the smallest PTVp volume (average 129cc), while an offline correction on the bony anatomy results inthe largest PTVp volume (average 192 cc). The PTVPLN volume issmallest for online correction on bony anatomy and largest foronline correction on prostate fiducials (average 606 cc versus940 cc). The average volume of PTVtotal varied between 762 ccwhen performing an online correction on bony anatomy and1108 cc when using the standard margin of 1 cm. No statisticallysignificant differences were observed in the coverage and meandose to the PTVs, ensuring that any OAR sparing is not caused bydifferences in PTV dose.

Lymph node margins (cm)

P LR SI AP

.1 0.5 0.5 0.7

.2 0.8 0.7 0.90.6 1.0 1.2

.0 1.0 1.0 1.0

adiation therapy strategy for organs at risk sparing in radiotherapy of the/10.1016/j.radonc.2018.02.009

Fig. 1. Visualisation of CTVs and PTVs for different correction strategies (light blue: CTVPLN and CTVSV, blue: CTVp, red: PTV with the standard margin of 1 cm. (a): PTV forcorrection strategy A (b): PTV for strategy B, (c): PTV for strategy C).

Table 2Dose criteria for the treatment plans.

PTVsPTVp V95% � 95%PTVPLN V95% � 95%PTVtotal Dmax < 107%

OARsRectum Dmax,0.5cc < 70 Gy

V69 Gy < 5%V66 Gy < 15%V58 Gy < 40%V30 Gy < 75%

Anal canal Dmean < 36 GySmall bowel bag Dmax,2cc < 52 Gy

V45 Gy < 195 ccFemoral heads Dmax,0.5cc < 50 Gy

Table 3Patient characteristics.

Number of patients n = 30

Age (years) 69 (57–77)Gleason score6/7/8/9/10 1/10/8/9/2PSA (ng/ml) 20 (2–70)T stage1/2/3/4 2/8/19/1N stage0/1 0/30M stage0/1 30/0

A. van Nunen et al. / Radiotherapy and Oncology xxx (2018) xxx–xxx 3

The evaluated parameters for rectum and anal canal for the dif-ferent strategies are visualised in Fig. 2. For all parameters exceptthe volume receiving 69 Gy, an online correction on the prostateleads to the largest reduction in rectum dose. No statistically sig-nificant differences were found in the mean rectum dose for theother strategies (averages varying between 37.5 Gy and 37.9 Gy).For strategy C, the average mean rectum dose was 31.9 Gy.

Please cite this article in press as: van Nunen A et al. Optimal image guided rprostate including pelvic lymph nodes. Radiother Oncol (2018), https://doi.org

Strategy B generally leads to the highest rectum dose, except forthe low dose region (V30Gy). The mean anal canal dose was lowestwhen applying strategy C and highest for strategy B (average 16.1Gy versus 23.5 Gy). The largest dose reduction for the rectum andanal canal was achieved by performing strategy C, online correc-tion on prostate fiducials.

The lowest mean bladder dose is achieved for strategies A and C(average 44.8 Gy and 44.3 Gy) and the highest dose for strategies Band D (average 46.3 Gy and 46.2 Gy), as displayed in Fig. 2.

adiation therapy strategy for organs at risk sparing in radiotherapy of the/10.1016/j.radonc.2018.02.009

Table 4Summary treatment planning results: dose and volume parameters in PTVs and OAR for evaluated strategies.

M: mean; SD: Standard Deviation; MDN: median; Min: minimum value; Max: maximum value.*p < 0.001 for all six pairwise comparisons.#p < 0.05 for all six pairwise comparisons.a–e: p > 0.05 in subgroups (when a pairwise comparison was not significant, this pair is noted by for example a – a in superscript).Green value represents the significant lowest value of all evaluated strategies.

4 Radiotherapy for prostate and lymph nodes

Strategy A results in the lowest SBB dose and strategies C and Dresult in the highest dose. The median maximum dose in 0.5 cc ofthe SBB was 49.9 Gy for strategy A and 50.8 Gy – 50.7 Gy for strate-gies C and D. The median absolute volume of the SBB whichreceived 45 Gy was 11 cc for strategy A and 18 cc for strategies Cand D. The largest dose reduction for the small bowel bag wasachieved by performing strategy A.

Compared to the standard 1 cm PTV margin, all strategies leadto a statistically significant reduction in the maximum dose tothe femoral heads. Strategy A results in the lowest maximum doseto the femoral heads, with an average maximum dose of 36.6 Gyfor the left femur head and 36.3 Gy for the right femur head.

Discussion

The aim of this study was to quantify the OAR dose for differentposition correction strategies and their corresponding PTV mar-gins, and to determine which strategy is most optimal for treatingpatients with lymph node positive prostate cancer.

Please cite this article in press as: van Nunen A et al. Optimal image guided rprostate including pelvic lymph nodes. Radiother Oncol (2018), https://doi.org

PTV volumes depend on the setup correction strategy as thisdetermines the required CTV-PTV margins. Online correction onprostate fiducials (strategy C) results in the smallest PTVp volumeand a larger PTVtotal volume due to larger margins around thePLN to account for misalignments between prostate and PLN. Thisleads to a larger overlap between PTVPLN and the bladder, femoralheads and SBB. Online correction on bony anatomy (strategy A)results in the smallest PTVtotal volume, but in an increase in thePTVp volume. This results in a larger overlap between PTVp andthe bladder, rectum and anal canal. Similar results were found byKershaw et al. [9]. They concluded that neither correction strategywas optimal, and that the dosimetric consequences should be fur-ther investigated. As different dose levels are prescribed to theprostate and PLN, it is difficult to directly translate the differencesin overlap to their influence on the OAR dose.

Although online correction on the prostate fiducials is resultingin a higher PTVtotal volume compared to correction on PLN, themost optimal sparing of the clinically relevant OARs can beachieved with this strategy. The smaller overlap between the PTVp

adiation therapy strategy for organs at risk sparing in radiotherapy of the/10.1016/j.radonc.2018.02.009

Fig. 2. Evaluated doses for rectum, bladder and anal canal.

A. van Nunen et al. / Radiotherapy and Oncology xxx (2018) xxx–xxx 5

volume and the OARs results in a decrease in the mean rectumand anal canal dose, as well as a reduction in all evaluatedDVH-parameters for the rectum. The mean bladder dose wasreduced compared to the offline correction strategies and wassimilar compared to the online correction strategy on bony anat-omy. Although the larger overlap between PTVtotal and the bladderfor strategy C, apparently the increased overlap of PTVPLN iscompensated by the (smaller) decrease in overlap with PTVp.When applying an online correction on prostate fiducials, the doseto the SBB and femoral heads was slightly higher compared tocorrection on bony anatomy due to the larger margins neededfor the PLN to account for baseline shifts. If the SBB dose isviolated in a patient, an online correction on PLN could bechosen to limit this dose at the cost of a higher rectum and analcanal dose.

Differences in OAR dose for online versus offline correction onbony anatomy were marginal. The random setup errors of thePLN are usually limited, while the larger variations in prostate

Please cite this article in press as: van Nunen A et al. Optimal image guided rprostate including pelvic lymph nodes. Radiother Oncol (2018), https://doi.org

position are not correlated to the bony anatomy. As a result, anonline correction on bony anatomy only results in a small reduc-tion of the PTV margins compared to an offline correction. An off-line strategy could thus be chosen when only correcting on thebony anatomy to reduce workload and imaging dose to the patient.Online correction on prostate fiducials leads to a much largerreduction in OAR dose and should thus be preferred.

Several studies investigated the effect of different correctionstrategies on target coverage [1,3,10,11,14]. These studies focussedon evaluating the target dose for a given PTV margin, reportingonly the target dose without evaluating the OAR doses. The mar-gins used in these studies varied from 5 mm to 10 mm and couldsometimes vary from patient to patient depending on the prefer-ence of the radiation oncologist [1,3]. In our study, anisotropicCTV-PTV margins based on an analysis of pre-treatment CBCTimages of 20 patients were used [6]. Compared to the other studies,the applied margins were slightly larger in AP direction, and some-what smaller in the other directions. Hsu et al., Thörnqvist et al.

adiation therapy strategy for organs at risk sparing in radiotherapy of the/10.1016/j.radonc.2018.02.009

6 Radiotherapy for prostate and lymph nodes

and Rossi et al. concluded that outliers (specifically due to changesin rectal filling) cause a deterioration of the CTV coverage, but formost patients their margins were adequate. The applied marginsin this study will take care that the CTV coverage is not compro-mised, even when these outliers occur during the radiotherapytreatment. Using standard CTV-PTV margins of 1 cm might notalways result in an optimal CTV coverage as the required setupmargins in anterior–posterior (AP) direction must be larger than1 cm [6]. These margins furthermore do not result in the lowestOAR dose.

Although there are no studies comparing the OAR doses fordifferent correction strategies, a number of authors have reportedon OAR dose for this patient group for a fixed PTV margin. Ishiiet al. made a dosimetric comparison between different treatmenttechniques (IMRT and VMAT) for patients with high-risk prostatecancer [15]. The margins around the prostate in this study wereslightly smaller than those in strategy C, and also smaller marginswere used around the PLN. A slightly higher dose was prescribed toPTVp (70 Gy compared to 67.2 Gy) while the dose to PTVPLN wasidentical. When comparing the OAR dose for strategy C of the cur-rent study with the VMAT results of Ishii et al., the mean bladderdose was marginally higher in the present study (44.3 vs 43.1Gy). A much lower mean rectum dose was however achieved(31.9 vs 37.5 Gy by Ishii et al.), and also V65-66 Gy was reduced.Additionally, the V45Gy of the SBB is lower in the present study(30 cc compared to 170 cc found by Ishii et al.). Although the PTVvolumes in our study were larger, similar or lower OAR doses wereachieved compared to Ishii et al. This is probably due to the slightlyhigher prescribed dose to the prostate, but also to differences in theoptimisation strategy, resulting in a more optimal sparing of theOARs in our plans.

McCammon et al. investigated the OAR dose and associated tox-icity for 30 patients [16]. They used the same fractionation schemeas Ishii et al. and also report higher OAR doses while using a 5 mmCTV-to-PTV margin around the PLN. Reported doses to the bladder,rectum and SBB were even higher than those in the study of Ishiiet al. For the rectum, a mean dose of 47.4 Gy was found comparedto 37.9 and 31.9 Gy for the least and most optimal strategy in thisstudy, respectively. For the SBB, V45Gy was not evaluated, but thereported volume receiving more than 50 Gy (64 cc) was higherthan the V45Gy in this study. From a toxicity analysis, McCammonet al. concluded that this treatment was well tolerated with lowtoxicity rates. With the reported doses in the present study,toxicity rates should be even lower.

This study has some limitations. The OAR dose was only evalu-ated using a single fractionation scheme and treatment technique.Choosing a different fractionation or technique might result in dif-ferent OAR doses and possibly another optimal strategy. Addition-ally, the use of a 6D couch for setup correction including rotationswas not investigated. A 6D correction strategy might result insmaller treatment margins and a better sparing of OARs. The dosecomparison was performed based on the margins needed for thedifferent strategies, and evaluated on the planning CT. Althoughthis is still common practice in radiotherapy, a more accurate esti-mation of the OAR dose could be obtained by including the dailyanatomical changes visible on the CBCTs of these patients. By recal-culating the treatment plans on the CBCT images and accumulatingthe dose, a better estimation of the actually delivered dose can beobtained.

As was also mentioned by Rossi et al., the margins needed forsufficient CTV coverage are to a large extent determined by outlierscaused by changes in rectal filling [1]. An adaptive strategy mightbe suitable to deal with these outliers, enabling a further reduction

Please cite this article in press as: van Nunen A et al. Optimal image guided rprostate including pelvic lymph nodes. Radiother Oncol (2018), https://doi.org

in the PTV margins, and consequently, a further reduction intoxicity. An offline adaptive strategy for example can help in deal-ing with outliers due to systematic error in the prostate position onthe planning CT. Further research is needed to quantify the reduc-tion in OAR dose using adaptive strategies.

In conclusion, although the total PTV volume is higher whenapplying an online correction strategy on the prostate due to thelarger margins around the PLN, this strategy leads to the most opti-mal sparing of the relevant OARs (rectum, anal canal and bladder),at the cost of a slightly higher dose to the femoral heads and smallbowel bag compared to the other correction strategies. Onlinecorrection on prostate fiducials using CBCT imaging is now usedclinically in our hospital.

Conflict of interest

Regarding our manuscript "Optimal image guided radiationtherapy strategy for organs at risk sparing in radiotherapy of theprostate including pelvic lymph nodes", none of the authors hasany conflict of interest to declare.

References

[1] Rossi PJ, Schreibmann E, Jani AB, Master VA, Johnstone PAS. Boost first,eliminate systematic error, and individualize CTV to PTV margin when treatinglymph nodes in high-risk prostate cancer. Radiother Oncol 2009;90:353–8.

[2] Meijer GJ, de Klerk J, Bzdusek K, et al. What CTV-to-PTV margins should beapplied for prostate irradiation? Four-dimensional quantitative assessmentusing model-based deformable image registration techniques. Int J RadiatOncol Biol Phys 2008;72:1416–25.

[3] Hinton BK, Fiveash JB, Wu X, Dobelbower MC, Kim RY, Jacob R. Optimalplanning target volume margins for elective pelvic lymphatic radiotherapy inhigh-risk prostate cancer patients. ISRN Oncol 2013;2013.

[4] Michalski JM, Gay H, Jackson A, Tucker SL, Deasy JO. Radiation dose-volumeeffects in radiation-induced rectal injury. Int J Radiat Oncol Biol Phys 2010;76.

[5] Kavanagh BD, Pan CC, Dawson LA, et al. Radiation dose-volume effects in thestomach and small bowel. Int J Radiat Oncol Biol Phys 2010;76:S101–7.

[6] van Nunen A, Budiharto T, de Vocht B, Schuring D. Evaluation of setup marginsusing cone-beam CT for prostate and pelvic nodes irradiation. Radiother Oncol2017;123:571.

[7] Van Herk M. Errors and margins in radiotherapy. Semin Radiat Oncol2004;14:52–64.

[8] Van Herk M, Remeijer P, Rasch C, Lebesque JV. The probability of correct targetdosage: Dose-population histograms for deriving treatment margins inradiotherapy. Int J Radiat Oncol Biol Phys 2000;47:1121–35.

[9] Kershaw L, van Zadelhoff L, Heemsbergen W, Pos F, van Herk M. Image guidedradiation therapy strategies for pelvic lymph node irradiation in high-riskprostate cancer: motion and margins. Int J Radiat Oncol Biol Phys2018;100:68–77.

[10] Hsu A, Pawlicki T, Luxton G, Hara W, King CR. A study of image-guidedintensity-modulated radiotherapy with fiducials for localized prostate cancerincluding pelvic lymph nodes. Int J Radiat Oncol Biol Phys 2007;68:898–902.

[11] Eminowicz G, Dean C, Shoffren O, Macdougall N, Wells P, Muirhead R.Intensity-modulated radiotherapy (IMRT) to prostate and pelvic nodes: ispelvic lymph node coverage adequate with fiducial-based image-guidedradiotherapy? Br J Radiol 2014;87:20130696.

[12] Lawton CAF, Michalski J, El-Naqa I, et al. RTOG GU radiation oncologyspecialists reach consensus on pelvic lymph node volumes for high-riskprostate cancer. Int J Radiat Oncol Biol Phys 2009;74:383–7.

[13] Bel A, van Herk M, Bartelink H, Lebesque JV. A verification procedure toimprove patient set-up accuracy using portal images. Radiother Oncol1993;29:253–60.

[14] Thörnqvist S, Hysing LB, Zolnay AG, et al. Treatment simulations with astatistical deformable motion model to evaluate margins for multiple targetsin radiotherapy for high-risk prostate cancer. Radiother Oncol2013;109:344–9.

[15] Ishii K, Ogino R, Okada W, Nakahara R, Kawamorita R, Nakajima T. Adosimetric comparison of RapidArc and IMRT with hypofractionatedsimultaneous integrated boost to the prostate for treatment of prostatecancer. Br J Radiol 2013;86:20130199.

[16] McCammon R, Rusthoven KE, Kavanagh B, Newell S, Newman F, Raben D.Toxicity assessment of pelvic intensity-modulated radiotherapy withhypofractionated simultaneous integrated boost to prostate for intermediate-and high-risk prostate cancer. Int J Radiat Oncol Biol Phys 2009;75:413–20.

adiation therapy strategy for organs at risk sparing in radiotherapy of the/10.1016/j.radonc.2018.02.009