Brachytherapy - (2014) -

Evaluation of intrafraction motion of the organs at risk in image-basedbrachytherapy of cervical cancer

Vijai Simha1,*, Firuza Darius Patel1, Suresh Chander Sharma1, Bhavana Rai1,Arun Singh Oinam1, Rahul krishnatry2, Bhaswanth Dhanireddy1

1Department of Radiotherapy and Oncology, Post Graduate Institute of Medical Education and Research (PGIMER), Chandigarh, India2Department of Radiation Oncology, Tata Memorial Hospital, Mumbai, India

ABSTRACT PURPOSE/INTRODUCTION: To assess the

Received 21 Janu

accepted 12 May 201

Grants: This study

Conflicts of intere

* Corresponding

Post Graduate Institu

Sector 12, Chandigarh

E-mail address: v

1538-4721/$ - see fro

http://dx.doi.org/10

variation in the doses received by the organs at risk(OARs) that can occur during treatment planning of cervical cancer by image-based brachytherapy.METHODS AND MATERIALS: After intracavitary application, two sets of imagesdCT andMRIdwere obtained. The two sets of images were fused together with respect to the applicator.Contouring was done separately on CT and MR images. Dose received by the OARs on CT imageswith respect to the plans made on the MR images was estimated and compared with those on theMR images.RESULTS: Although there was always a difference between the dose received by the OARs basedon the CT and MRI contours, it was not significant for the bladder and rectum; 2 cc doses differedby 0.49 Gy (�0.44) p5 0.28 for the bladder and 0.30 Gy (�0.29) p5 0.16 for the rectum. The 1 ccand 0.1 cc differences were also not significant. However for the sigmoid colon, there was signif-icant intrafraction variation in the 2 cc doses 0.61 (�0.6) p 5 0.001, 1 cc doses 0.73 (�0.67) Gyp 5 0.00, and 0.1 cc dose 0.97 (�0.93) Gy p 5 0.009.CONCLUSIONS: The variation in the doses to the OARs must be considered while weighingtarget coverage against overdose to the OARs. Although not significant for the bladder and rectum,it was significant for the sigmoid colon. Estimated doses to OARs on the planning system may notbe the same dose delivered at the time of treatment. � 2014 American Brachytherapy Society. Pub-lished by Elsevier Inc. All rights reserved.

Keywords: Image-based brachytherapy; Organs at risk; Organ motion; Optimization; Intrafraction motion; Intrafraction

changes

Introduction

Brachytherapy in cervical cancer plays a very importantrole in obtaining high cure rates with minimum complica-tions. Traditionally, the doses delivered to the organs at risk(OARs) during cervical brachytherapy have beenrepresented by International Commission on RadiationUnits and Measurement points (1). It is increasingly being

ary 2014; received in revised form 1 May 2014;

4.

did not require any grants.

st: None to report.

author. Department of Radiotherapy and Oncology,

te of Medical Education and Research (PGIMER),

160012, India. Tel.: þ91-991-417-9639.

ijaiaditya1985@gmail.com (V. Simha).

nt matter � 2014 American Brachytherapy Society. Publis

.1016/j.brachy.2014.05.016

recognized that the dose to the OARs cannot be representedby points as there is a steep dose gradient because of theproximity to the brachytherapy sources (2, 3). With theintroduction of CT, the external contours of the bladderand rectum can be determined with a reasonable accuracyand the organ can be delineated as a single volume (4, 5).Sectional image-based planning enables greater accuracyand reproducibility of topography of the OARs and reliablecalculation of the doses received by the OARs (5, 6). CTand MRI provide equal quality for discrimination of thebladder, rectum, and sigmoid (7). Delineation of organ con-tour permits a reliable estimation of the dose actually deliv-ered to 2 and 1 cc of the wall of the organ exposed to a highdose. The parameters D2cc, D1cc, and D0.1cc (doses receivedby most exposed part of 2, 1, and 0.1 cc volume, respec-tively) for bladder, rectum, and sigmoid colon are usuallyrecorded (8). Studies have demonstrated doseeeffect

hed by Elsevier Inc. All rights reserved.

2 V. Simha et al. / Brachytherapy - (2014) -

relationship to small volumes of OARs irradiated and havefound good correlation between the dose to 2 cc volumesand Grade 2e4 toxicity (9, 10).

We would like to qualify the term intrafraction motion.In teletherapy, it would be applicable to the motion occur-ring during the short time the patient is actually undergoingtreatment. However in brachytherapy, it is important torealize that the changes that occur after implantation ofthe applicator and before and during the treatment alsohave an impact on the final treatment plan, optimization,and dose distribution, and hence these events are includedunder the term fraction. Image-based brachytherapy is along labor intense procedure, and there is a definite possi-bility that changes in the volume and position of the OARscan occur. These differences in volumes and geometry ofthe OARs can arise from the differences in the filling andmotion of the OARs, the movement of the patient duringtransfer of the patient from one place to another, or alsoto a lesser extent from interobserver differences in contour-ing and OAR movement with respiration (11). The netresult from these various motions is that there may be a dif-ference in the doses calculated and the doses actuallyreceived by the OARs. The practical way to substantiatethis difference is to measure the change in the dosesreceived by small volumes of OARs with respect to theapplicator, which will ultimately carry the radioactivesource. Despite the ability of the CT scans to quantifythe doses received by small volumes of the OARs, thereproducibility of the same doses in the patient duringthe actual treatment is also critical as nowadays, moreand more emphasis is laid on optimization to limit thedoses to the OARs (12, 13). If there are significant differ-ences in the doses calculated and doses received, then thewhole practice of optimization in image-based brachyther-apy may be called into question. Also as the intermediaterisk clinical target volume (IRCTV) is limited by the OARs(14), the exact delineation of the IRCTV is dependent onthe OAR geometry and may change as it changes withtime.

Methods and materials

Between November 2010 and January 2013, we treated50 patients with our in-house image-based brachytherapyprotocol, which consisted of a pretreatment MRI, chemora-diation, followed by MR image-based brachytherapy.External radiation was delivered by three-dimensionalconformal radiotherapy to the whole pelvis, and a dose of46 Gy in 23 fractions was delivered over a period of 4 ½weeks with weekly cisplatin at a dose of 40 mg/m2.Brachytherapy was delivered in four fractions of 7 Gy each.During brachytherapy, applicator was inserted undergeneral anesthesia and immobilized with compact gauzepacking of the vagina and a stay suture on the vulva. AT-bandage was also used to secure the applicator to the

pelvis. An MRI was done for identification of the targetvolumes. Patients also underwent a CT scan to facilitatethe planning process as it helped in applicator reconstruc-tion as our department did not have the library model plansfor reconstruction on Oncentra Master Plan version 3.0software (Nucletron, an Elekta company, Elekta AB, Stock-holm, Sweden). There was a time gap of an average of 2 h(1/2e3½ h) between acquisition of MRI and CT andaverage 7 h (5e8 h) between applicator placement andtreatment delivery with the major delay being in acquiringthe MRI as it had been done in the radiology department. Abladder-filling protocol was followed before acquisition ofthe CT and MRI and also before treatment delivery in anattempt to produce uniform bladder filling. Fifty millilitersof normal saline was introduced into the bladder throughthe Foleys catheter and allowed to drain naturally by grav-ity. MRI was acquired on Siemens Magnetom 3 T (SiemensAG, Munich, Germany) with sections acquired at 3-mmintervals keeping distance factor zero. CT was acquiredon GE LightSpeed CT Scanner (GE Healthcare, ChalfontSt. Giles, UK, a unit of General Electric Company) withsections acquired at 2.5 mm thickness. After image acqui-sition, primary delineation of the target and OAR delinea-tion was done on the MRI. The MR images were fusedwith the CT images using the manual registration methodof Oncentra (Elekta). The applicator geometries in three-dimensional MR image were reconstructed manuallyaccording to CT image, fused on MR image using the rigidregistration of applicator geometry to an accuracy of�1 mm. CTeMRI fusion was done with respect to theapplicator geometry only. Contouring was done accordingto Groupe Europ�een de Curieth�erapie and the EuropeanSociety for Radiotherapy & Oncology recommendations(14), and the outer wall of the OARs was contoured. Doseof 7 Gy was prescribed to the high-risk clinical target vol-ume (HRCTV). Contouring on both the sets of images wasdone by only one observer (VS) to eliminate interobservervariations. Plan optimization was done to improve thetarget coverage or to reduce the doses to the OARs. Patientswere treated with a microselectron high-dose rate (Nucle-tron) unit in the brachytherapy suite. Retrospectively, con-touring of the OARs was done on CT scan, and the dosesreceived by the OARs as delineated on the CT was evalu-ated for the treatment plan used to treat with MRI. Analysisof the difference in the doses of the paired data was doneusing the Wilcoxon signed rank test.

Results

In each of the 50 patients evaluated in this study, someamount of OAR movement always occurred betweenacquisition of CT and MRI. No contour of any of theOAR was exactly geometrically similar when the MR andCT images were superimposed on each other (Figure 1).The average change in the volume of the bladder between

Fig. 1. The difference in the position of the organs at risk that can occur in relation to the applicator in the period between the acquisition of CT and MRI.

The outer contours of the bladder and rectum show minimal change with respect to the applicator, whereas the sigmoid can show gross changes because of its

free mobility.

3V. Simha et al. / Brachytherapy - (2014) -

the CT and MRI was 15.7 (�13.8) cc and for the rectum itwas 7.8 (�6.7) cc (Table 1). Similarly, change in the vol-ume for the sigmoid was 14.9 (�13.25) cc. The changein the geometry and volume of the OARs also resulted inthe changes of the D2cc, D1cc, and D0.1cc doses. The averagechange in the 2 cc doses of the bladder between CT andMRI was 0.49 (�0.44) Gy, 0.30 (�0.29) Gy for the rectum,and 0.61 (�0.60) Gy (Table 2) for the sigmoid. Similarvariations were noted in the D0.1cc, D1cc, doses as well(Table 2). With respect to our prescribed dose of 7 Gyper fraction, 11 of the 50 patients had a variation in thebladder dose of greater than 10% (0.7 Gy). Variation ofgreater than 10% of the prescribed dose was seen in 6 pa-tients for the rectum and 18 patients for the sigmoid. Tosummarize the dosimetric differences, the variationsoccurred most for the sigmoid followed by the bladderand least for the rectum (Figure 2).

In all the patients after superimposing MRI onto theCT, it was seen that the IRCTV contoured on the MRimages did not hold good for the CT images because ofmotion and geometric displacement of the OARs (Figs.1 and 2).

Table1

The average and median volumes of the OARs in MR and CT images, the avera

OAR change

in volume Mean (�SD) cc Median (IQR)

Bladder 15.7 (�13.8) 12.27 (�12.18)

Rectum 7.8 (�6.7) 6.43 (�7.85)

Sigmoid 14.9 (�13.25) 9.62 (�15)

OARs 5 organs at risk; SD 5 standard deviation; IQR 5 interquartile range

Discussion

In this study, we have attempted to define dosimetric dif-ferences because of variations and displacement of theOARs in relation to the applicator. For OARs, the minimumdose in the most irradiated tissue volume is recommendedfor reporting: 0.1, 1, and 2 cc (8). Doseeeffect relationshipsfor induction of late side effects have been proposed for theOARs (9, 10). For the urinary bladder, D2cc is considered,for which equivalent dose at 2 Gy per fraction (EQD2) of100 Gy can be used as clinical cutoff for morbidity. Forrectal morbidity, D2cc related with Grade 2e4 side effectsand a dose of 78 Gy EQD2 resulted in a 10% probabilityof Grade 2 side effects (10). However for the sigmoid, alow probability of side effects was observed (10), perhapsbecause of high variability of sigmoid topography betweenapplications. In an observational study by Sturdza et al.(15), of 22 patients, not a single case could be identifiedin which sigmoid loops were identically positioned in theproximity of the applicator between the fractions. In theabsence of definite data regarding the sigmoid toxicityand the observed intrafraction variations observed in thepresent study, we recommend that dose must not be

ge changes in the volumes, and the range of variation in the volumes

cc

Minimum change

in volume (cc)

Maximum change

in volume (cc)

0.22 61.45

0 28.72

0.66 52.79

.

Table 2

The average doses to the 2, 1, and 0.1 cc of the bladder, rectum, and sigmoid and the range of the changes in the dose that can occur between the two sets of

images taken on the same day

OAR dose

MRI dose CT dose Difference in MRI and CT doses

p-Value

Range of difference

Mean (�SD) Gy Maximumeminimum Gy

Bladder (cc dose)

2 5.52 (�1.29) 5.65 (�1.1) 0.49 (�0.44) 0.28 0.01e1.911 6.08 (�1.37) 6.23 (�1.19) 0.56 (�0.55) 0.32 0.2e2.47

0.1 7.4 (�1.59) 7.68 (�1.49) 0.84 (�0.74) 0.26 0.06e4.29

Rectum (cc dose)

2 2.72 (�1.29) 2.88 (�0.91) 0.30 (�0.29) 0.16 0e1.261 3.01 (�1.37) 3.22 (�1.04) 0.42 (�0.59) 0.48 0e3.13

0.1 3.59 (�1.59) 3.91 (�1.29) 0.50 (�0.66) 0.12 0.1e3.56

Sigmoid (cc dose)

2 4.75 (�1.29) 4.35 (�1.37) 0.61 (�0.6) 0.001 0e2.41

1 5.35 (�1.37) 4.86 (�1.53) 0.73 (�0.67) 0.000 0e2.75

0.1 6.53 (�1.59) 6.07 (�2.2) 0.97 (�0.93) 0.009 0e3.98

SD 5 standard deviation.

4 V. Simha et al. / Brachytherapy - (2014) -

compromised to the HRCTV to keep the sigmoid doseswithin limits. In our study, the doses to the HRCTV andIRCTV was assumed to be the same on the CT and MRIas the applicator was rendered immobile in relation to cer-vix by use of compact vaginal gauze packing and also byuse of a labial stay suture. Though variations were foundbetween the D2cc, D1cc, and D0.1cc for the bladder andrectum, but they were not statistically significant. However,the sigmoid colon has a broad fan-shaped mesentery with along attachment at its origin allowing for varying degrees ofdistension and pressure according to its luminal contents.This extreme mobility conferred on the sigmoid colonmay even result in a volvulus of that bowel (16). In ourstudy, as a result of the mobility of the sigmoid, there

Fig. 2. A fused sagittal view of CT and MRI with respect to the applicator. The

similar relation to the applicator. However, the sigmoid colon shows significant

was a significant difference in the dose to the 2, 1, and0.1 cc of the sigmoid between CT and MR imaging.Although the 2 cc doses showed an average change ofnearly 10% of the prescribed dose, the variation in the 1and 0.1 cc doses was much higher than 10%. Although sta-tistically we have shown that there is not much dosimetricintrafraction variation for the bladder and rectum, large dif-ferences in the D2cc doses have occurred in individualpatients in this study. Change in D2cc dose as high as1.91 Gy (27.2% change with respect to prescribed dose)for the bladder 1.26 Gy (18% change with respect to pre-scribed dose) for the rectum and 2.41 Gy (34.4% changewith respect to prescribed dose) for the sigmoid have beenseen. In a similar study by Anderson et al. (17), a

posterior wall of the bladder and the anterior wall of the rectum maintain a

change in position owing to its free mobility.

5V. Simha et al. / Brachytherapy - (2014) -

pretreatment MRI was taken after an average of 4.75 h ofplanning MRI in 21 patients, and it was seen that althoughthe mean changes in the dose to the OARs was not signif-icant, significant changes in individual patients did occur.Also the time elapsed between the two MRI scans in thatstudy did not correlate with the change in the OAR doses.Our study differs from the study by Anderson in that thevariations in the doses to the sigmoid were significant. Itmust be further noted that this variation depicted in ourstudy is only for a single fraction and if similar variationis produced in multiple fractions (our protocol administersfour fractions of 7 Gy at brachytherapy); the final effect onthe uncertainty of the EQD2 dose received by the OARs islikely to get further magnified.

Whenever a plan is optimized so as to decrease the doseto the OARs, it always means constricting of the isodosecurves around the applicator and a fall in the HRCTV andPoint A dose. Dramatically optimizing standard treatmentplans so as to decrease the dose to the OAR is not justifiedas we have shown here that there may be a significantchange in the dose between two sets of images taken onthe same day. Also it is well accepted that a poorly opti-mized plan is worse than a standard plan where the pre-scribed dose is normalized to Point A. Also as inferredfrom this study, the geometry and volume of IRCTV (asdefined by Groupe Europ�een de Curieth�erapie and the Eu-ropean Society for Radiotherapy & Oncology) (8) is likelyto change as the geometry of OARs changes with time.Thus, the IRCTV is best regarded as a conceptual ratherthan an actual entity, and the margins given around theHRCTV to contour the IRCTV should not be limited bythe OARs as the topography of the OARs is likely tochange over a period, and sometimes the OARs may welllie adjacent to the HRCTV.

The limitations of this study are that the variations in theOAR doses have been evaluated only between CT and MRimaging and not between planning and treatment delivery.Although it would be advisable to keep the time betweenplanning and treatment to the minimum, it must be pointedout that the changes can occur to increase and decrease thevariation onto the next day as seen in a study by Lang et al.(18). While estimating the total dose to an OAR as EQD2,it has always been assumed in the past that the same area ofthe OAR is in the high-dose region, although this may notbe true as there is organ motion and deformation betweenthe fractions and also within a fraction as seen in this study.Also it is possible that applicator displacement (if inade-quately secured) can occur between image acquisitionand treatment. This adds another element of uncertaintyin the practice of image-based brachytherapy.

Summary

This study quantifies the volume changes that occur inthe OARs and the resultant changes in the doses to small

volumes of the OAR that can occur because of volumeand position changes. Despite intrafractional variations,our results encourage the practice of cautious optimizationin image-based brachytherapy considering the fact thatdose calculated may not be the exact dose that is receivedby the OARs but is likely to be very close as far as thebladder and rectum are concerned. For the sigmoid colon,this study has shown that there is likely to be significant dif-ferences in the dose received over a period as it is freelymobile within the pelvis. Hence, overenthusiastic optimiza-tion to decrease the dose to the sigmoid colon is not encour-aged. Also occasionally, dramatic variation of the doses canoccur for the bladder and rectum on an individual basis, theclinician must thoroughly justify the consequent fall in thedose to HRCTV that can occur because of optimization.The variation in the topography of the OARs may explainsome of the unexplained toxicities that do not correlate withthe doses delivered. Hence, it is important to realize thatplanned dose to the OAR is not necessarily the dosereceived by these organs at the time of treatment delivery.

References

[1] Chassagne D, Dutreix A, Almond P. ICRU report no. 38: Dose and

volume specification for reporting intracavitary therapy in gynecol-

ogy. Bethesda, MD: ICRU; 1985.

[2] Lahtinen T, Tenhunen M, Vayrynen M. ICRU reference points and

maximum doses of rectum and bladder in intracavitary radiotherapy.

Radiother Oncol 1993;28:174e176.

[3] Visser AG, Symonds RP. Dose and volume specification for reporting

gynaecological brachytherapy: Time for a change. Radiother Oncol

2001;58:1e4.

[4] Wachter-Gerstner N, Wachter S, Reinstadler E, et al. The impact of

sectional imaging on dose escalation in endocavitary HDR-

brachytherapy of cervical cancer: Results of a prospective compara-

tive trial. Radiother Oncol 2003;68:51e59.

[5] Wachter-Gerstner N, Wachter S, Reinstadler E, et al. Bladder and

rectum dose defined from MRI based treatment planning for cervix

cancer brachytherapy: Comparison of dose-volume histograms for or-

gan contours and organ wall, comparison with ICRU rectum and

bladder reference point. Radiother Oncol 2003;68:269e276.

[6] Krishnatry R, Patel FD, Singh P, et al. CT or MRI for image-based

brachytherapy in cervical cancer. Jpn J Clin Oncol 2012;42:

309e313.

[7] Viswanathan AN, Dimopoulos J, Kirisits C, et al. Computed tomog-

raphy versus magnetic resonance imaging-based contouring in cervi-

cal cancer brachytherapy: Results of a prospective trial and

preliminary guidelines for standardized contours. Int J Radiat Oncol

Biol Phys 2007;68:491e498.

[8] Potter R, Haie-Meder C, Van Limbergen E, et al. Recommendations

from gynaecological (GYN) GEC ESTRO working group (II): Con-

cepts and terms in 3D image-based treatment planning in cervix can-

cer brachytherapy-3D dose volume parameters and aspects of 3D

image-based anatomy, radiation physics, radiobiology. Radiother On-

col 2006;78:67e77.

[9] Georg P, Lang S, Dimopoulos JC, et al. Dose-volume histogram pa-

rameters and late side effects in magnetic resonance image-guided

adaptive cervical cancer brachytherapy. Int J Radiat Oncol Biol Phys

2011;79:356e362.

[10] Georg P, Potter R, Georg D, et al. Dose effect relationship for late

side effects of the rectum and urinary bladder in magnetic resonance

6 V. Simha et al. / Brachytherapy - (2014) -

image-guided adaptive cervix cancer brachytherapy. Int J Radiat On-

col Biol Phys 2012;82:653e657.

[11] Prabhakar R, Jagadesan P, Haresh KP, et al. A study on the variation

of bladder and rectal doses with respiration in intracavitary brachy-

therapy for cervix cancer. J Contemp Brachytherapy 2010;2:24e27.[12] Potter R, Dimopoulos J, Georg P, et al. Clinical impact of MRI assis-

ted dose volume adaptation and dose escalation in brachytherapy of

locally advanced cervix cancer. Radiother Oncol 2007;83:148e155.

[13] P€otter R, Dimopoulos J, Georg P. Impact of systematic MRI assisted

3D treatment planning on local control and morbidity in cervix can-

cer: Vienna experience in 125 patients treated by intracavitary þ/-

interstitial HDR brachytherapy from 1998e2003. Radiother Oncol

2006;81:S109.

[14] Haie-Meder C, Potter R, Van Limbergen E, et al. Recommendations

from Gynaecological (GYN) GEC-ESTRO Working Group (I):

Concepts and terms in 3D image based 3D treatment planning in cer-

vix cancer brachytherapy with emphasis on MRI assessment of GTV

and CTV. Radiother Oncol 2005;74:235e245.

[15] Sturdza AE, Berger D, Lang S, et al. Uncertainties in assessing sig-

moid dose volume parameters in MRI-guided fractionated HDR

brachytherapy. Brachytherapy 2008;7:109.

[16] Snell R. Clinical anatomy by regions. Philadelphia, PA: Wolters

Kluwer Health; 2011.

[17] Anderson C, Lowe G, Wills R, et al. Critical structure movement in

cervix brachytherapy. Radiother Oncol 2013;107:39e45.

[18] Lang S, Nesvacil N, Kirisits C, et al. Uncertainty analysis for 3D

image-based cervix cancer brachytherapy by repetitive MR imaging:

Assessment of DVH-variations between two HDR fractions within

one applicator insertion and their clinical relevance. Radiother Oncol

2013;107:26e31.