t

L

Medical Dosimetry 37 (2012) 163-169

Medical Dosimetry

journal homepage: www.meddos.org

d

Can radiation therapy treatment planning system accurately predict surface dosesin postmastectomy radiation therapy patients?

Sharon Wong, M.Sc.,* Michael Back, M.D.,† Poh Wee Tan, B.Sc.,‡ Khai Mun Lee, M.D.,‡

Shaun Baggarley, B.Sc.,‡ and Jaide Jay Lu, M.D.*‡

*National University of Singapore, Yong Loo Lin School of Medicine, Singapore; †Northern Sydney Cancer Centre, Royal North Shore Hospital, St Leonards, New South Wales, Australia; and‡National University, Cancer Institute, Department of Radiation Oncology, National University, Hospital, Tower Block, Singapore

A R T I C L E I N F O

Article history:

A B S T R A C T

Skin doses have been an important factor in the dose prescription for breast radiotherapy. Recent advances

Received 29 September 2010Accepted 10 June 2011

in radiotherapy treatment techniques, such as intensity-modulated radiation therapy (IMRT) and newtreatment schemes such as hypofractionated breast therapy have made the precise determination of thesurface dose necessary. Detailed information of the dose at various depths of the skin is also critical in

designing new treatment strategies. The purpose of this work was to assess the accuracy of surface dosecalculation by a clinically used treatment planning system and those measured by thermoluminescencedosimeters (TLDs) in a customized chest wall phantom. This study involved the construction of a chest wallphantom for skin dose assessment. Seven TLDs were distributed throughout each right chest wall phantomto give adequate representation of measured radiation doses. Point doses from the CMS Xio� treatmentplanning system (TPS) were calculated for each relevant TLD positions and results correlated. There wereno significant difference between measured absorbed dose by TLD and calculated doses by the TPS (p �

0.05 (1-tailed). Dose accuracy of up to 2.21% was found. The deviations from the calculated absorbed doseswere overall larger (3.4%) when wedges and bolus were used. 3D radiotherapy TPS is a useful and accuratetool to assess the accuracy of surface dose. Our studies have shown that radiation treatment accuracyexpressed as a comparison between calculated doses (by TPS) and measured doses (by TLD dosimetry) canbe accurately predicted for tangential treatment of the chest wall after mastectomy.

� 2012 American Association of Medical Dosimetrists.

Keywords:Breast phantomDosimetryTangential irradiationAbsorbed dosesTreatment planning system

d

Introduction

Local recurrence in the chest wall is a commonmode of treatmentfailure in patients with locally advanced breast cancer after mastec-tomy, and occurs in 3–22% of postmastectomy breast cancer patientsdepending on the size, location, and estrogen receptor (ER) status ofthe primary disease, presence of regional lymph nodemetastasis, anduse of adjuvant treatment.1,2 Skin overlying the tumor bed and mas-ectomy scar are at the highest risk of local recurrence.3

Radiation therapy is a commonly usedmodality for adjuvant treat-ment in patients with locally advanced breast cancer after mastec-tomy. Results from randomized trials have demonstrated that post-mastectomy radiotherapy improves prognoses, including overallsurvival in patients with locally advanced breast cancer.4,5 It is gener-ally accepted that the improvement in treatment outcome, including

Reprint requests to: Jaide Jay Lu, M.D., National University of Singapore, Yong Loo

in School of Medicine, 21 Lower Kent Ridge, Road, Singapore 119077.

E-mail:mdcljj@nus.edu.sg

0958-3947/$ – see front matter Copyright � 2012 American Association of Medical Dosimetroi:10.1016/j.meddos.2011.06.006

disease-free survival (DFS) and overall survival (OS) is secondary to, atleast in part, improved local control.

Skin doses have always been an important factor in the dose pre-scription for postmastectomy radiotherapy. Radiation dose to the su-perficial part of the chest wall, including the skin after mastectomy isassociated with the probability of local control.6,7 Prospective ran-omized studies3,6–8 have demonstrated the effectiveness of a skin

boost in reducing the risk of local recurrence, evenwith negativemar-gins after postmastectomy irradiation. Sufficient surface dose canthus improve local control and minimize local recurrence. However,skin toxicity secondary to radiotherapy is one of the most frequentlyobserved adverse effects after postmastectomy irradiation. Althoughreversible in most cases, skin toxicity can affect the therapeutic strat-egy and adverse effect patients’ quality of life.9

Dose at the surface of the radiation field, such as skin in the post-mastectomy cases is primarily caused by electron contamination.10,11

It is affected by field size, angle of beam incidence, air gap, and the useof beam modifiers.12 Accurate skin dose assessment is particularlychallenging in the postmastectomy chest wall because of the tangen-

tial beams and the large curvature of the chest and the limited chest

ists

n

ra

olTt

ba

tTS

umt

T

rpur1B1

T

tpfcde

S. Wong et al. / Medical Dosimetry 37 (2012) 163-169164

wall thickness. Inaccuracy of calculations in the build-up region willalso affect the doses given to the normal skin tissues and for potentialdisease control near skin surface. Because both local disease controland the prevalence and severity of skin toxicity are radiation dose–related, knowledge of skin dose is clearly important in identifyingareaswhere unwanted skin reactions can be expected and in ensuringthat structures at risk of recurrence receive the prescribed dose. Al-though the accuracy of skin dose measurement by commerciallyavailable radiation planning systems has been confirmed in irradia-tion for head-and-neck disease,13 such accuracy on the chest wall hasot been confirmed in postmastectomy radiation therapy.Thermoluminescence dosimetry (TLD) measurement is a proven

eliable method for the verification of many of the dosimetric aspectsssociated with external beam radiotherapy.14–17 It is the gold stan-

dard dosimetry program recommended for quality assurance of ma-chine calibration, planning dosimetry, and dose calculation. The pur-pose of this study was to assess the accuracy of surface dosecalculation in postmastectomy radiation by comparing the surfacedose predicted by a clinically used treatment planning system (TPS)and those measured by TLDs in a customized chest wall phantom.

Methods and materials

Construction of mastectomy phantom

This study involves the construction of a chest wall phantom for skin dose assess-ment. Most commercially available radiation therapy anthropomorphic phantomshave very thick chest wall measuring at least �5 cm. Thus it cannot be used to accu-rately provide skin dose measurement for postmastectomy patients in whom chestwall thickness was measured to range from 2–5 cm only.18

The phantom is constructed of materials designed to conform to the requirementsf ICRUReport Number 44. The densities of the structures in the phantom,whichmodelung tissue, have a mass density of 0.32 g/cm3, which is similar to normal lung density.

Fig. 1. Chest wall thickness and lung measured using CT transversal slice (A). Thephantom was then constructed according to the curvature of the lung and chest wallusing wax and cork material. A customized chest wall phantom using wax and cordmaterial (B).

he entire phantom was manufactured using wax from a mold made from a templateaken from the transversal computed tomography (CT) slices of a mastectomy patient.

The selection of wax material was used, because it has similar mass density of normaltissue. Thephantom is capable of having thermoluminescent dosimeter (TLD)materialsinserted at various locations to enable dose assessment. Chest wall thickness and lungvolume was measured using CT transversal slice of a postmastectomy patient (Fig. 1).

Materials and Methods

TLD placements and in vivo dosimetry

Seven TLDs were distributed throughout each right breast phantom to give ade-quate representation of dose distribution, including the superior and inferior planesrelative to the central axis (z� 0), where greater dose inhomogeneity is known to occur(superior and inferior planes used were z � � 6 cm and –6 cm).

These 7 positions were chosen to avoid regions of potential dose gradient and toreflect the dose delivered to the surface of the chest wall. Each plane has 3 separate TLDpositions (Fig. 2). The TLDs were placed in customized cavities that were individuallymolded to fit the TLD chip thickness at the surface and entrance dose region. Surfacedose is here defined as the dosemeasured at a distance�5mmbelow the patient’s skin,and entrance dose region is measured at a distance (dmax; 1.5 cm for the 6-MV beam)elow the patient’s surface. To assess the reproducibility of the measurements, thebsorbed dose measurement with TLDs was repeated 5 times for each region.

A display of TLD positions for the phantom are presented in Figs 3 and 4. Note thathe customized phantomhas a thin chest wall and a crescent-shaped 2D breast outline.LD measurements were repeated with placement of 1-cm bolus over the chest wall.imilarly, 7 TLDs were placed under the bolus.

The customized mastectomy phantom underwent CT scanning for dose planningsing the same procedures as those for a patient. In this particular case, a set of scans (3m) was acquired, which made dose calculations in any cross-section through the

arget volume possible.

LDs

The TLD chips used for the study were hot-pressed LiF chips (3.1 � 3.1 � 0.9 mm)ead on an automated TLD reader (Realto Automatic TLD 1A Processor, Nuclear Enter-rises, Edinburgh, UK) and annealed in a TLD oven. The TLD chips underwent regularniformity measurements to monitor reproducibility and estimate each chip’s relativeesponse. The annealing protocol was 1 hour of heating at 400�C followed by 2 hours at00�C. The chips were then read by the automated reader and annealed again for reuse.efore commencing the study, the TLD chips were individually calibrated in groups of0 and were kept in the same group to ensure each TLD was as identical as possible.

PS

A CMS XiO� 3D TPS (Elekta CMS Software, Inc., Stockholm, Sweden) was used forhis investigation. All treatment fields included lateral tangential fields using 6-MVhotons, and a virtual wedge of 10� was used for contour compensation. Point dosesrom the Xio� TPS were calculated for each relevant TLD position. The use of lungorrectionwas taken into account. Surface and build-up dose valueswere calculated byetermining the peak dose along the central axis of each defined field using point-dosestimates from the planning system. For comparison with possible off-axis positioning

Fig. 2. Display of TLD positions on the mastectomy phantom.

a2

S

wcTm

s

entati

S. Wong et al. / Medical Dosimetry 37 (2012) 163-169 165

of TLDs, surface and build-up dose values were also calculated at off-axis locations asdescribed before.

The convolution/superposition algorithmwas used for the tangential breast calcu-lations. Treatment plans were generated using 8 � 16-cm2 tangential fields at beamngles of 50� and 230� to simulate a chest wall treatment for a prescribed dose of00 cGy.

tatistical analysis

Statistical Package for Social Sciences, version 17.0 (SPSS, Inc., Chicago, IL)as used for data analysis. The measurements of each of the 7 TLD readings wereompared with corresponding points obtained by TPS by using a 1-sample t-test.his test determined the significance of differences between the values of the sameeasurements made under the 2 different conditions (measured doses by TLDs

Fig. 3. Schematic repres

Fig. 4. Schematic representation of the TLD po

vs. calculated doses by TPS). A p-value � 0.05 (1-tailed) was consideredignificant.

Results

All values are a representation of the combined contribution fromboth medial and lateral tangential beams. The distribution of the de-viations between TLD and calculated radiation dose was considered.The measurements were also analyzed using wedges and the pres-ence of bolus.

The results of the measurements are presented in Tables 1 and 2and Figs 5 and 6. Tables 1 and 2 show the surface doses between

on of the TLD positions.

sitions on a phantom with 1-cm bolus.

Table 1The mean surface dose measurements of all 7 positions as measured in the customized mastectomy breast phantom

Surface dose measurements

Plans without bolus or wedge (control) Plan with 1-cm bolus Plan with wedge Plan with bolus and wedge

Slices through thebreast volume

TLD(Gy)

TPS(Gy)

Mean diff (%) andp sig if � 0.05*

TLD(Gy)

TPS(Gy)

Mean diff (%) andp sig if � 0.05*

TLD(Gy)

TPS(Gy)

Mean diff (%) andp sig if � 0.05*

TLD(Gy)

TPS(Gy)

Mean diff (%) andp sig if � 0.05*

Superior medial regionz � 60 mm

1.52 1.50 1.20 (p � 0.183) 1.91 1.95 �2.09 (p � 0.181) 1.56 1.55 0.72 (p � 0.213) 1.91 1.92 �0.44 (p � 0.36)

Central axis medialregion z � 0 mm

1.66 1.65 0.77 (p � 0.308) 1.93 1.95 �0.57 (p � 0.305) 1.61 1.64 �1.56 (p � 0.156) 1.96 1.95 0.44 (p � 0.11)

Inferior medial regionz � -60 mm

1.95 1.94 0.36 (p � 0.230) 2.06 2.10 �1.88 (p � 0.307) 1.91 1.94 �1.63 (p � 0.129) 2.14 2.10 2.01 (p � 0.09)

Central axis isocenterregion Z � 0 mm

1.87 1.89 �1.36 (p � 0.103) 2.15 2.15 �0.20 (p � 392) 1.90 1.94 1.64 (p � 0.130) 2.10 2.10 �0.23 (p � 0.42)

Superior lateral regionz � 60 mm

1.83 1.84 �0.50 (p � 0.351) 1.94 1.95 �0.42 (p � 0.390) 1.94 1.93 0.65 (p � 0.202) 1.98 2.00 �0.94 (p � 0.27)

Central axis lateralregion z � 0 mm

1.84 1.80 2.21 (p � 0.078) 2.07 2.07 0.18 (p � 0.455) 1.90 1.9 0.19 (p � 0.467) 2.10 2.08 1.00 (p � 0.27)

Inferior lateral regionZ � �60 mm

1.83 1.80 1.70 (p � 0.178) 2.06 2.07 �0.55 (p � 0.255) 1.86 1.92 �3.15 (p � 0.157) 2.13 2.12 0.34 (p � 0.30)

The doses relate to 2 Gy given as mean target dose in the breast.* A single-sample test was performed on SPSS ver 17.

Table 2The mean entrance dose (build-up region) measurements of all 7 positions as measured in the customized mastectomy breast phantom

Entrance dose(buildup region) measurements

Plans without bolus or wedge (control) Plan with 1-cm bolus Plan with wedge Plan with bolus and wedge

Slices through thebreast volume

TLD(Gy)

TPS(Gy)

Mean diff (%) andp sig if � 0.05*

TLD(Gy)

TPS(Gy)

Mean diff (%) andp sig if � 0.05*

TLD(Gy)

TPS(Gy)

Mean diff (%) andp sig if � 0.05*

TLD(Gy)

TPS(Gy)

Mean diff (%) andp sig if � 0.05*

Superior medial regionz � 60 mm

1.94 1.93 0.4 (p � 0.218) 1.98 1.99 �0.49 (p � 0.409) 1.98 1.98 0.435 (p � 0.256) 2.05 2.03 0.94 (p � 0.172)

Central axis medialregion z � 0 mm

1.95 1.99 �1.85 (p � 0.155) 2.00 2.03 �1.64 (p � 0.250) 1.97 1.98 �0.413 (p � 0.233) 2.06 2.08 �0.93 (p � 0.193)

Inferior medial regionz � -60 mm

1.99 1.99 0.03 (p � 0.485) 2.13 2.12 0.29 (p � 0.210) 2.08 2.10 �0.753 (p � 0.189) 2.15 2.14 0.82 (p � 0.279)

Central axis isocenterregion Z � 0 mm

1.92 1.97 �2.56 (p � 0.103) 1.97 2.00 �1.35 (p � 0.110) 1.93 1.95 1.002 (p � 0.079) 1.95 1.98 �1.38 (p � 0.430)

Superior lateral regionz � 60 mm

1.94 1.96 �0.73 (p � 0.253) 2.04 2.09 �2.34 (p � 0.161) 2.07 2.05 1.002 (p � 0.192) 2.07 2.10 �1.54 (p � 0.119)

Central axis lateralregion z � 0 mm

1.99 2.02 �1.31 (p � 0.069) 2.14 2.15 �0.29 (p � 0.422) 2.08 2.09 �0.008 (p � 0.498) 2.18 2.19 �0.30 (p � 0.101)

Inferior lateral regionz � �60 mm

1.95 1.97 �1.22 (p � 0.195) 2.06 2.07 �0.61 (p � 0.449) 2.10 2.09 0.476 (p � 0.235) 2.23 2.20 1.54 (p � 0.194)

The doses relate to 2 Gy given as mean target dose in the breast.* A single-sample test was performed on SPSS ver 17.

S.Wong

etal./M

edicalDosim

etry37

(2012)163-169

166

sSnhs

dmt

a

tvs

sc

S. Wong et al. / Medical Dosimetry 37 (2012) 163-169 167

absorbed dose (TLD) and calculated doses (TPS) in different slicesthrough the mastectomy breast phantom. In general, there is no sig-nificant difference between measured absorbed dose and calculateddoses by the TPS (p� 0.05 (1-tailed). Dose accuracy of up to 2.21%wasobtained. The deviations from the calculated absorbed doses are over-all larger (3.4%) when wedges and bolus are used.

With 1-cm bolus, significant skin dose build-up was seen. Table 1hows the surface dose without bolus and with 1-cm bolus material.urface dose without bolus is much smaller than that with bolus. Theet result is that at 1-cm bolus, skin doses are approximately 39%igher at surface. The calculated skin dose at similar points showimilar results, although readings were slightly higher (45%).

Table 2 shows the entrance dose at the build-up region (1.5-cmepth) between absorbed dose (TLD) and calculated doses (TPS) in theastectomy breast phantom. Results show lesser dose variation in

he dmax. region, with variations between calculated and measuredreadings in themain cross-section of the breast reduced to 2.56% only.

The measured absorbed doses in the whole target volume are ingeneral lower than calculated levels and the differences are largerwhen wedges and bolus are used as beam modifiers and compensa-tors. The measured dose distribution also shows a smaller variationthan the calculated one.

Discussion

The objective of this study was to investigate whether it was pos-sible to use radiation therapy TPS to accurately predict the radiation

0

0.5

1

1.5

2

2.5

SupMed

CAXmed

InfMed

CAXIso+

Sup Lat CAX lat Inf Lat

TLD vs TPS surface dose

TLD (control)

TPS (control)

TLD (bolus)

TPS (bolus)

0

0.5

1

1.5

2

SupMed

CAXmed

Inf Med CAXIso+

Sup Lat CAX lat Inf Lat

TLD (control)

TPS (control)

TLD (wedge)

TPS (wedge)

0

0.5

1

1.5

2

2.5

SupMed

CAXmed

InfMed

CAXIso+

SupLat

CAXlat

InfLat

TLD (control)

TPS (control)

TLD (bolus+wedge)

TPS (bolus+wedge)

A

B

C

Fig. 5. (A) TLD vs. TPS surface dose (build-up) with 1-cm bolus; (B) TLD vs. TPSurface dose (build-up) with wedge; (C) TLD vs. TPS surface dose (build-up) with 1-m bolus and wedge.

skin dose in postmastectomy radiotherapy patients. The calculatedd

radiation skin dose by TPS was compared with actual measured doseby 7 small TLDS placed in a customized mastectomy phantom. Al-though results show some degree of variations between these 2 dosemeasurements, however, the discrepancies in our study are small,with dose variations up to 2.56%. This is in good agreement with arecent study byMijnheer et al.,19 who recommend a 2–3.5% degree ofccuracy. Redpath et al.20 carried out a similar study in a range of

situations using geometrical phantoms and also obtained an agree-ment of 2–3%.

However when compared with similar studies done on breast re-gions, our studies demonstrated different results. Our results differedfrom studies byWestermann et al.21 and Knoos et al.,22 in which theyreported that tangentially applied fields may give an overestimationof the dose calculation of up to 8%. This can be easily explained be-cause of current improved 3D treatment planning algorithms andfaster computers compared with those used in studies done in the1980s.

Skin toxicity is one of the most frequent adverse reactions thatoccur in patients who undergo postmastectomy radiotherapy. Theresults obtained from this study are important because accurate sur-face dose and entrance dose determination allows the sequentialchecking of thedosimetry aspects over thewhole treatment chain, theassessment of the quality of a given irradiation technique, and thecontrol of the dose delivery in individuals.23 The ability to estimatehe surface and superficial dose with radiation therapy TPSs can pro-ide valuable information for clinical consideration to avoid near-urface recurrence while at the same time limiting severe skin toxic-

1.81.85

1.91.95

22.05

2.12.15

SupMed

InfMed

SupLat

InfLat

TLD vs TPS entrance dose

TLD (control)

TPS (control)

TLD (bolus)

TPS (bolus)

1.8

1.85

1.9

1.95

2

2.05

2.1

SupMed

CAXmed

InfMed

CAXIso+

SupLat

CAXlat

InfLat

TLD (control)

TPS (control)

TLD (wedge)

TPS (wedge)

1.7

1.8

1.9

2

2.1

2.2

2.3

SupMed

CAXmed

InfMed

CAXIso+

SupLat

CAXlat

InfLat

TLD (control)

TPS (control)

TLD (bolus +wedge)

TPS (bolus + wedge)

A

B

C

Fig. 6. (A) TLD vs. TPS entrance dose (dmax) with 1-cm bolus; (B) TLD vs. TPS entrance

ose (dmax) with wedge; (C) TLD vs. TPS entrance dose (dmax) with 1-cm bolus and

wedge.

btu

aaatpt

mtTedala

aitsdpspf

a

S. Wong et al. / Medical Dosimetry 37 (2012) 163-169168

ity, especially for postmastectomy chest wall treatments. Asdiscussed, studies have showed that the skin overlying the tumor bedand mastectomy scar are the regions that are at the greatest risk forrecurrence, with most chest wall failures occurring in these regions.3

A separate study by Huang et al.8 have shown improved locoregionalcontrol in postmastectomy patients when sufficient skin dose is de-livered.

Improved surface dose prediction can also improve local controland minimize local recurrence.1 If these individuals could be recog-nized in advance of treatment, their clinical radiotherapy regimescould be adjusted accordingly to optimize results, and this will beeffective in achieving a significant improvement in quality and out-come of breast treatments in mastectomy patients.

Many studies have focused on the skin dose for head and necktumors. For example, Lee et al.24 identified severe skin reactions insome patients receiving intensity-modulated radiation therapy(IMRT) for head andneck cancer. They found that the average increaseof the skin dose was about 18% caused by the bolus effect of a ther-moplasticmaskmaterial and the skin doseswithmaskswere 84% and100% of the prescribed dose for opposed lateral and extended field.They concluded that potential contributions to the increased skindose are the bolus effect of the head and neck and shoulder mask, theuse of multiple oblique beams, and the inclusion of the skin as part ofthe target volume. Higgins et al.13 investigated delivery techniquedifferences and found that the maximum surface doses of head andneck deliverywas different between bilateral fields, tomotherapy andIMRT, respectively. On the basis of previous studies, many factorsaffect the doses of the skin and near-surface tumor, e.g., the use ofbolus, oblique incident angle, and delivery technique. Although fac-tors affecting the skin dose for head and neck treatments have beenstudied, little is known about that for chest wall radiotherapy.

For chest wall tangential radiotherapy, 1-cm bolus is often usedduring the treatment course to ensure adequate dose to the targetvolume, which includes the skin, but it may be removed if deemedclinically necessary when skin toxicity occurs. Our study evaluatedthe effect of 1-cm bolus on the surface dose as well. The surface dosewith boluswas determined to bemuch larger than thatwithout bolus,as expected (39% increase in dose when bolus was used). The calcu-lated skin dose by TPS at similar points shows similar results, althoughreadings were slightly higher (45%). In comparison, a recent study byQuach et al.25 found that the surface dose with 1-cm bolus increasedy 54% compared with the dose without bolus on a chest wall phan-om. Because of the different thickness and curvature of the phantomsed, our increase rateswere different fromQuach et al.25 The effect of

the bolus material on the surface dose is thus dependent on the dif-ferent geometry and thickness of the chest wall.

As can be observed, the treatment geometry caused by the tangen-tial beams, the use of wedges, and the large curvature of the chestmake surface dose prediction difficult. However, our results and re-cent studies by Spenzi et al.26 and Quach et al.25 have shown thatccurate clinical dose calculation can be computed by radiation ther-py TPSs, provided that the radiation source and the patient anatomyre modeled correctly. This explains the use of a customized mastec-omy phantom in our study tomodel the exact treatment geometry ofostmastectomy breast patients to produce a higher degree of radia-ion treatment planning accuracy.

It is important to recognize the limitations of this research. Theain limitation of our study is the use of only 7 TLD positions placed

hroughout the mastectomy phantom. Ideally, a larger number ofLDs placed throughout the phantomwould bemore desirable; how-ver, this was not necessary because these 7 positions were evenlyistributed throughout the chest wall; thus accurate interpolation ofny points within the chest wall can be made to reflect the dose de-ivered to the skin. These 7 positions were also carefully chosen to

void regions of potential dose gradient and scatter. The ability to

ccurately position the TLDs on the curved phantom’s surface for eachrradiationwas also a challenge. Accurate TLD placement is an impor-ant factor in determining measured dose. A study by Herbert16 hashown that TLD positioning errorswould lead to changes inmeasuredose as a result of changes in source-to-surface distance caused byatient contour, changes in wedge transmission, and variation in re-ponse of the TLDs. To avoid TLD placement error, each set of TLDositions was repeated 5 times. Average of the 5 readings was takenor analysis.

Radiation skin dose of postmastectomy patients could be moreccurately studied in vivo in patients instead of with the use of a

mastectomy phantom. However, the phantom model used is con-structed with radiologically equivalent material for both soft tissueand lungs. It is equipped with holes in grid form, which providesaccess for the TLD chips. These holes allowed for measurements to betaken in the build-up region, as well as at various depths of the chestwall, whichwould not be possible in vivo in postmastectomy patients.The mastectomy phantom was specially constructed for this experi-ment becausemost commercially available radiation therapy anthro-pomorphic phantoms have a very thick chest wall measuring at least�5 cm. Therefore, it cannot be used to accurately provide skin dosemeasurement for postmastectomy patients whose chest wall thick-ness wasmeasured to range from 2–5 cm only.18 Finally, because thisstudy was done entirely with the use of a phantom model, actualpatients’ movement and breathing was excluded. As such, we are notsure whether our results would differ with actual radiation dosesmeasured in vivo in patients.

Conclusion and Future Direction

The goal of this research was to provide clinicians informationabout the accuracy of surface dose prediction by radiation therapyplanning systems in postmastectomy breast patients. This is of greatimportance to obtain knowledge of skin dose to identify areas whereunwanted skin reaction would be expected or to ensure that struc-tures at risk of recurrence receive their prescribed doses. Our studieshave shown that radiation treatment accuracy expressed as a com-parison between calculated doses (by TPS) and measured doses (byTLD) can be accurately predicted for tangential treatment of the chestwall after mastectomy.

Because we have studied and validated our radiation therapy TPSas an accurate prediction tool for surface dose measurements, an im-mediate application would be to evaluate clinical skin reactions inpostmastectomy patients. Future work includes the correlation ofclinical data in postmastectomy patients with predicted doses by ra-diation therapy planning systems to allow better evaluation of theprobability of undesirable effects of radiotherapy. Thesemeasures canlead to prediction of skin reactions, helping with the design of newtreatment techniques and different dose fractionation schemes. If thedesign of new treatment techniques can aim to control the dose tovarious points according to likelihood of recurrence, this can greatlyimprove local control and disease-free survival in patients using post-mastectomy radiotherapy.

References

1. Buchanan, C.L.; Dorn, P.L.; Fey, J.; et al. Locoregional recurrence after mastectomy:Incidence and Outcomes. J. Am. Coll. Surg. 203:469–74; 2006.

2. Magno, L.; Bignardi, M.; Micheletti, E.; et al. Analysis of prognostic factors in pa-tientswith isolated chestwall recurrence of breast cancer. Cancer 60:240–4; 1987.

3. Donegan, W.L.; Perez, M.C., Watson, F.R. A biostatistical study of locally recurrentbreast carcinoma. Surg. Gynecol. Obstet. 122:529–40; 1966.

4. Overgaard, M. Overview of randomized trials in high risk breast cancer patientstreated with adjuvant systemic therapy with or without postmastectomy irradia-tion. Semin. Radiat. Oncol. 9:292–9; 1999.

5. Ragaz, J.; Jackson, S.M., Le, N.; et al. Adjuvant radiotherapy and chemotherapy in

node-positive premenopausal womenwith breast cancer.N. Engl. J. Med. 337:956–62; 1997.

1

1

1

1

1

1

1

1

1

1

2

2

2

2

2

2

S. Wong et al. / Medical Dosimetry 37 (2012) 163-169 169

6. Wong, J.S. Postmastectomy radiation therapy. In: Diseases of the breast.Harris, J.R.Boston: Lippincott Williams &Wilkins; 2004:785–99.

7. Bentzen, S.M., Overgaard, M. Relationship between early and late normal tissueinjury after postmastectomy radiotherapy. Radiother. Oncol. 20:159–65; 1991.

8. Huang, E.Y.; Chen, H.C.; Sun, L.M.; et al.Multivariate analyses of locoregional recur-rences and skin complications after postmastectomy radiotherapy using electronsor photons. Int. J. Radiat. Oncol. Biol. Phys. 65:1389–96; 2006.

9. Pires, A.M.T.; Segreto, R.A., Segreto, H.R. RTOG criteria to evaluate acute skin reac-tion and its risk factors in patients with breast cancer submitted to radiotherapy.Rev. Lat. Enfermagem. 16:844–9; 2008.

0. Klein, E.E., Purdy, J.A. Entrance and exit dose regions for a Clinac-2100C. Int. J.Radiat. Oncol. Biol. Phys. 27:429–35; 1993.

1. Parthasaradhi, K.; Prasad, S.G.; Rao, B.M.; et al. Investigation on the reduction ofelectron contamination with a 6-MV x-ray beam.Med. Phys. 16:123–5; 1989.

2. Hsu, S.H.; Roberson, P.L.; Chen, Y.; et al. Assessment of skin dose for breast chestwall radiotherapy as a function of bolus material. Phys. Med. Biol. 53:2593–606;2008.

3. Higgins, P.D.;Han, E.Y.; Yuan, J.L.; et al.Evaluation of surface and superficial dose forhead and neck treatments using conventional or intensity-modulated techniques.Phys. Med. Biol. 52:1135–46; 2007.

4. Essers, M., Mijnheer, B. In vivo dosimetry during external photon beam radiother-apy. Int. J. Radiat. Oncol. Biol. Phys. 43:245–59; 1999.

5. Herbert, C.E.; Ebert,M.A., Joseph, D.J. Feasiblemeasurement errorswhen undertak-ing in vivo dosimetry during external beam radiotherapy of the breast.Med. Dosim.28:45–8; 2003.

6. Kron, T. Thermoluminescence dosimetry and its applications in medicine—part 1:physics, materials and equipment. Australas. Phys. Eng. Sci. Med. 17:175–82; 1994.

7. Lee, P.C.; Sawicka, J.M., Glasgow, G.P. Patient dosimetry quality assurance pro-gram with a commercial diode system. Int. J. Radiat. Oncol. Biol. Phys. 29:1175–82; 1994.

8. Ramsey, C.R.; Chase, D.; Scaperoth, D.; et al. Improved dose homogeneity to theintact breast using three-dimensional treatment planning. Med. Dosim. 25:1–6;1998.

9. Heukelom, S.; Lanson, J.H.; van Tienhoven, G.; et al. In vivo dosimetry during tan-gential breast treatment. Radiother. Oncol. 22:269–79; 1991.

0. Redpath, A.T.; Thwaites, D.I.; Rodger, A.; et al. A multidisciplinary approach toimproving the quality of tangential chest wall and breast irradiation for carcinomaof the breast. Radiother. Oncol. 23:118–26; 1992.

1. Westermann, C.F.;Mijnheer, B.J., VanKleffens, H.J. Determination of the accuracy ofdifferent computer planning systems for treatment with external photon beams.Radiother. Oncol. 1:339–47; 1984.

2. Kn××s, T.; Ahlgren, L., Nilsson,M. Comparison ofmeasured and calculated absorbeddoses from tangential irradiation of the breast. Radiother. Oncol. 7:81–8; 1986.

3. Cozzi, L., Cozzi, A.F. Quality assurance in radiation oncology. A study of feasibilityand impact on action levels of an in vivo dosimetry program during breast cancerirradiation. Radiother. Oncol. 47:29–36; 1998.

4. Lee, N.; Chuang, C.; Quivey, J.M.; et al. Skin toxicity due to intensity-modulatedradiotherapy for head-and-neck carcinoma. Int. J. Radiat. Oncol. Biol. Phys. 53:630–7; 2002.

5. Quach, K.Y.; Morales, J., Butson, M.J.; et al.Measurement of radiotherapy x-ray skindose on a chest wall phantom.Med. Phys. 27:1676–81; 2000.

26. Spezi, E., Lewis, G. An overview of Monte Carlo treatment planning for radiother-apy. Radiat. Protect. Dosim. 131:123–9; 2008.