Enhanced epidermal dose caused by localized electron contaminationfrom lead cutouts used in kilovoltage radiotherapy

J. E. Lye,a� D. J. Butler, and D. V. WebbAustralian Radiation and Protection and Nuclear Safety Agency, Yallambie, Melbourne 3085, Australia

�Received 25 January 2010; revised 22 April 2010; accepted for publication 10 June 2010;published 12 July 2010�

Purpose: To investigate and quantify electron contamination from the lead cutouts used in kilo-voltage x-ray radiotherapy.Methods: The lead cutouts were modeled with the Monte Carlo EGSnrc user codes DOSXYZnrcand DOSRZnrc for x-ray beams ranging from 50 to 300 kVp. The results from the model wereconfirmed with Gafchromic film measurements. The model and measurements investigated the dosedistribution with and without gladwrap™ shielding under the lead, and dose distributions withround, square, and serrated edge cutouts.Results: Large dose enhancement near the edges of the lead was observed due to electron contami-nation. At the epidermal/dermal border, there is double the dose at the edge of the lead compared tothe central dose due to electron contamination for a 150 kVp beam and three times the dose for a300 kVp beam. gladwrap™ shielding effectively removes the contaminant dose enhancement usingten and four layers for 300 and 150 kVp beams, respectively.Conclusions: The contaminant dose enhancement is undesirable as it could cause unnecessaryerythema and hyperpigmentation at the border of the treated and untreated skin and lead to a poorercosmetic outcome. The contamination is easily removed by gladwrap™ shielding placed under oraround the lead cutout. © 2010 American Association of Physicists in Medicine.�DOI: 10.1118/1.3458722�

Key words: kilovoltage radiotherapy, electron contamination, lead cutout, skin dose

I. INTRODUCTION

The problem of electron contamination from the metal conesused to shape the x-ray beams in kilovoltage �kV� radio-therapy is an acknowledged complication in kV dosimetry.1

Protocols such as TRS-398 recommend using thin films toremove the electron contaminant dose when using parallelplate chambers to measure the x-ray output.2 The AAPMTG-61 kV protocol discusses the issue in some detail, notingthat the contamination may have clinical, as well as dosim-etric, ramifications, and should not be ignored duringtreatment.3 Electron contamination from lead eye shields isalso a known problem and commercial eye shields now havea protective low Z layer to soak up the contaminant dose.4

An issue that has been largely neglected to date is the ques-tion of electron contamination from the custom lead cutoutsthat are placed directly on the patient’s skin to define thetreatment area, as shown in Fig. 1. One study by Lee andChang5 considered this from a dosimetric point of view,measuring the central surface dose change with and withouta thin film electron filter on a parallel plate chamber. Theyobserved around a 10% change in surface dose for smalldiameter cutouts. A recent paper by Thomas and Clark6 in-vestigates using a thin coating painted onto the custom leadcutouts to reduce the electron dose, again measuring the ef-ficiency of this method with a thin window parallel platechamber.

The work presented in the current paper confronts the

problem from a new direction, using Monte Carlo �MC�

3935 Med. Phys. 37 „8…, August 2010 0094-2405/2010/37„8

tools to evaluate the contaminant dose over the whole treat-ment field, focusing particularly on the dose deposited lo-cally near the edge of the lead cutout. A very large localizedepidermal edge dose enhancement of up to 400% was dis-covered using MC modeling and Gafchromic film measure-ments. At a depth equivalent to the epidermal/dermal border,the edge dose enhancement remained high, up to 300%, andcould lead to unnecessary erythema and hyperpigmentationat the border of the treatment field.

II. METHODS

II.A. Monte Carlo modeling of the lead cutouts

The EGSnrc user codes DOSRZnrc and DOSXYZnrcwere used to model the dose from round and square leadcutouts, respectively. The MC transport parameters chosenwere those recommended for kV calculations in Ref. 7. TheDOSRZnrc and DOSXYZnrc simulations used 1�109 pri-mary histories and 1�108 primary histories, respectively,resulting in relative voxel dose accuracy of the order of 1%.No variance reduction was used in the simulations. Thesource was a photon point source. The beam radius at thesurface of the lead was 1 mm larger than the cutout underinvestigation. The source to surface distance was 30 cm. Thesource energies were the measured spectra of the AustralianRadiation Protection and Nuclear Safety Agency �AR-PANSA� kilovoltage radiotherapy calibration beams from aconstant potential, tungsten target, Seifert x-ray unit �Seifert

X-Ray Corp., Fairview Village, PA�. The spectra were mea-

3935…/3935/5/$30.00 © 2010 Am. Assoc. Phys. Med.

3936 Lye, Butler, and Webb: Enhanced epidermal dose from lead cutouts 3936

sured with a Compton spectrometer at 1.6 m from the tubeand showed good agreement with spectra obtained from asemiempirical model by Tucker.8 MC modeling of our kilo-voltage beam qualities have been validated in a previouswork.9 Two different geometries were used to simulate dosein a water phantom, and dose to film in air, as shown in Fig.1. For DOSRZnrc, 0.5 mm radius cylinders were used, andfor DOSXYZnrc, 0.5�0.5 mm2 voxel widths were used.The slabs of water shown in Fig. 1�a� were used for beamsfrom 150 to 300 kVp. Water slab thickness was reducedfrom 0.05 to 0.02 mm for 100 kVp and to 0.002 mm for50 kVp beams.

II.B. EBT film measurements

Gafchromic EBT film is well suited for superficial mea-surements as it does not require a light tight cardboard enve-lope that would absorb much of the contaminant electronsbefore they could be measured. The film was placed flushagainst the lead cutout at a distance of 30 cm from the x-raysource. Cones are not used at ARPANSA, but a 20 mm leadaperture 15 cm from the source constrains the beam diam-eter. The EBT has approximately 100 �m of Mylar before a50 �m active layer and therefore measures the dose depos-ited between 100 and 150 �m depth in Mylar. The EBT wascalibrated with the ARPANSA 300 kVp �HVL 3.8 mm Cu�beam. Known doses were delivered to ten film samples,which were then scanned after a period of 24 h, and a cali-bration curve of dose to scanner output was obtained. Thescanner was an Epson Perfection V700 photo scanner. Anoptical density dependant scanner nonuniformity correctionwas applied. Comparison of film profiles to profiles obtainedwith ionization chamber measurements indicate a relative

FIG. 1. Schematic of geometry used in DOSRZnrc and DOSXYZnrc for �a�modeling the dose in water and �b� modeling the dose to film. The model isrun with and without the polyethylene layer for each case. For DOSRZnrc0.5 mm radius, cylinders are used, and for DOSXYZnrc, 0.5�0.5 mm2

voxel widths are used.

film dose accuracy of the order of 1%.

Medical Physics, Vol. 37, No. 8, August 2010

III. RESULTS

III.A. Dose enhancement with and without gladwrap™

The first measurements were taken with the highest en-ergy 300 kVp beam incident on a 25 mm diameter 3 mmthick round lead cutout. Figure 2�a� shows images of dosedeposited between 100 and 150 �m depth in Mylar from thefilm measurement and the MC model. The dose profiles, nor-malized to the central dose, are shown below the images. Theclear dose enhancement near the edges of the lead perturbsthe dose distribution that ideally should be uniform. There isgood agreement between the model and the film measure-ments, with both showing a ratio of the edge dose to thecentral dose of approximately 1.4. The edge dose is slightlylower in places on the film compared to the MC in Fig. 2�a�where no gladwrap™ is used. This is presumably due to airgaps between the film and the lead as discussed in Sec. III C.

Figure 2�b� shows the same measurement when ten layersof gladwrap™ �approximately 200 �m of polyethylene� iswrapped around the lead. The polyethylene absorbs the elec-tron contamination and the desired uniformity in the dosedistribution is seen.

III.B. Dose enhancement with depth in a waterphantom

The dose enhancement was modeled with depth in a waterphantom using DOSXYZnrc to simulate a 20�20 mm2

square lead cutout with 300 �HVL 3.8 mm Cu� and 180 kVp

�HVL 0.99 mm Cu� beams. Figure 3 shows the profiles forfive different average depths in water from 35 to 235 �mand images showing the dose at three depths. The profiles aretaken across the center of the square and normalized to thecentral dose at a depth of 235 �m. The edge dose enhance-ment reduces rapidly with depth in water and reduces muchmore rapidly for the lower energy beam due to the reduced

FIG. 2. DOSRZnrc results and the film measurements of irradiation with a300 kV �HVL 3.8 mm Cu� beam, with and without ten layers of gladwrap™over the 25 mm diameter lead cutout.

penetration of the low energy electrons.

3937 Lye, Butler, and Webb: Enhanced epidermal dose from lead cutouts 3937

The images in Fig. 3 show that the edge enhancement ismuch higher in the corners of the square as opposed to thesides of the square. This suggests that the shape of the leadcutout is important in determining the edge dose enhance-ment. A variety of different shape and size cutouts weremodeled, including a serrated edge cutout or “pie crust” cut-out. Serrated edge cutouts are sometimes used to improve thecosmetic outcome by softening the residual scar border at theradiotherapy treatment site.10 An interesting question arisesas to whether the improved cosmetic result is due to theserrations smearing the edge dose over a larger area overmany fractions, or whether the serrated shape physically re-duces the local electron contamination. To address this point,the maximum and minimum edge dose enhancement in theserrated cutout is compared to the round and square cutouts.The image inset in Fig. 4 shows a film image of the dosefrom such a serrated edge cutout using the 300 kVp beam.The maximum edge dose enhancement occurs in the triangu-lar sections surrounded by the lead, while the minimum oc-curs in between the triangular sections. The serrated cutoutwas modeled by a 2 cm square with serrations spaced everytwo millimeters on one edge.

Figure 4 shows the change in edge dose enhancementwith depth in water for 1 and 2 cm diameter round cutouts,the side and corner dose enhancement for a 2 cm squarecutout, and the maximum and minimum edge dose enhance-

FIG. 3. Reduction of the contaminant edge dose with depth in water with nogladwrap™ from the MC model. A 300 and a 180 kV beam are considered.A 2�2 cm2 square lead cutout is investigated. The profiles show a sliceacross the center of the profile. Note that the contaminant edge dose ishigher in the corners of the square.

ment for the serrated lead cutout. For each series, the edge

Medical Physics, Vol. 37, No. 8, August 2010

dose is shown relative to the central dose at the maximumdepth of 235 �m. The minimum edge dose enhancement inthe serrated cutout was the same as the side dose enhance-ment in the square cutout. The maximum edge dose enhance-ment from the serrated cutout was much higher than thesquare cutout side dose enhancement. This suggests that theserrated shape does not physically reduce the electron con-taminant dose enhancement, but actually increases the localdose at the edge of the treatment field in a single fraction.The improved cosmetic outcome with a serrated edge may bedue to averaging out the edge dose over a larger area overmany fractions and the reduction of the hard line at the edgeof the treatment field. The serrated edge cutout was alsomodeled with 200 �m polyethylene under the lead. The ser-rated field shape is unchanged, but the ratio of edge dose-:central dose is reduced to less than 1.1 for all depths andedge positions.

A larger range of energies was modeled to detail the be-havior of the electron contaminant dose enhancement withdepth in water. A series of 2 cm square lead cutouts with anincident beam energy of 300 �HVL 3.8 mm Cu�, 250 �HVL2.5 mm Cu�, 180 �HVL 0.99 mm Cu�, 150 �HVL 0.60 mmCu�, 100 �HVL 6.53 mm Al�, and 50 kVp �HVL 0.79 mmAl� were modeled. An average of the side and corner doserelative to the central dose at 235 �m was taken as a broadindication of edge dose enhancement for all cutout shapes.Exponential fits to the edge dose enhancement versus depthin water is shown in Fig. 5 for all energies excluding 50 kV;50 kV was omitted for clarity as the dose enhancement didnot penetrate past the first 10 �m, equivalent to the deadouter layer of the epidermis.

The exponential fits to the edge dose enhancement havebeen overlayed on a schematic of the skin. The schematicshows the stratum corneum �the dead layer�, the basal layerof the epidermis at the epidermal/dermal border, and the up-

FIG. 4. The modeled dose enhancement at the edges of the lead cutout forround, square, and serrated edge cutouts, all for 300 kV HVL 0.38 mm Cu,from the MC model. There is no polyethylene shielding.

per dermis. The thickness of the epidermis is critical to un-

3938 Lye, Butler, and Webb: Enhanced epidermal dose from lead cutouts 3938

derstanding the problem posed by the shallow contaminantdose enhancement. The thickness of the epidermis variesgreatly with body site, as well as varying between individu-als. Whitton and Everall11 observed a forehead epidermisthickness of 50 �m versus a fingertip thickness of 369 �m.More recently, Gambichler et al.12 measured the average epi-dermal thickness on the forehead, pectoral, scapular, fore-arm, buttock, and calf to be approximately 70 �m for theyounger group and 60 �m for the older group. In this study,the results of Gambichler et al.12 are used. The foreheadthickness of the older group was taken as appropriate forkilovoltage radiotherapy, defining the epidermal thickness inthis study as 60 �m.

From Fig. 5 we can extract the edge dose enhancement atpertinent depths. Three depths were chosen as the most rel-evant for this study: The average epidermis dose, theepidermal/dermal border, and the upper dermis. Figure 6shows the edge dose enhancement as a function of x-ray tube

FIG. 5. The edge dose enhancement with depth in water for energies rangingfrom 300 to 100 kVp from the DOSXYZnrc model with no polyethyleneshielding. The edge dose enhancement is for a 2�2 cm2 square lead cutout,taking an average of the side and corner doses relative to the central dose at235 �m depth. An exponential fit of the MC data points is shown in thefigure. The figure is overlaid on a schematic of the structure of facial skin.

FIG. 6. The edge dose enhancement as a function of kV for three pertinent

depths. There is no polyethylene shielding in this model.

Medical Physics, Vol. 37, No. 8, August 2010

voltage for these depths. The epidermal/dermal border wasdefined as 60 �m depth. The average viable epidermal depthwas defined as 35 �m, excluding the first 10 �m of deadcells. The upper dermis was considered at a depth of100 �m. The epidermal/dermal border is radiobiologicallyimportant. Radiation damage at this depth can causeerythema due to inflammation from the breakdown of cells inthe actively dividing basal layer. Damage to the melanocytesfound in this layer can lead to melanin deposition in thedermis and result in local hyperpigmentation.13

Edge dose enhancement to the average epidermis is large;about four times the central dose at 235 �m and roughlyconstant from 150 to 300 kVp. This behavior can be quali-tatively understood by considering that although more con-taminant electrons will be produced from higher energy xrays, the electrons will also penetrate further and will notdeposit all of their energy at the shallow depths around35 �m. At the epidermal/dermal border, we see aboutdouble edge dose enhancement for 150 kVp increasing up tothree times edge dose enhancement for 300 kVp. For100 kVp and below, there is no significant edge dose en-hancement at the epidermal/dermal border depth. At the up-per dermis, 250 and 300 kVp show about double edge doseenhancement and below 150 kVp, there is no significantedge dose enhancement.

For each of the kilovoltage energies, layers of gladwrap™�20 �m polyethylene� were added to the model until thedose enhancement was reduced to less than 10% in the10–60 �m epidermal layer. Table I shows the results fromthis investigation.

III.C. Air gaps

The effect of an air gap between the lead cutout and theskin has been investigated. Figure 7 shows the case of a300 kVp beam incident on a 25 mm diameter lead cutoutmodeled with the lead flush on the water surface, with 0.2

FIG. 7. The MC dose profile for a round 2.5 cm lead cutout at a depthbetween 10 and 60 �m with no shielding, 0.2 mm of polyethylene, and 0.2

mm of air gap.

3939 Lye, Butler, and Webb: Enhanced epidermal dose from lead cutouts 3939

mm shielding of polyethylene, or with 0.2 mm of air gap.The profile is of the dose deposited between 10 and 60 �mwater depth, equivalent to the viable epidermis. The 0.2 mmof polyethylene cuts out the edge dose enhancement almostcompletely, while the air gap only reduces the contaminantdose slightly.

IV. CONCLUSIONS

The lead cutouts used in kilovoltage radiotherapy cause alarge electron contaminant dose enhancement at the edge ofthe treatment field at shallow depths. Such a dose enhance-ment is undesirable as it could lead to unnecessary erythemaand hyperpigmentation at the border of the treated and un-treated skin and a poorer cosmetic outcome. The edge doseenhancement at a particular depth in water is highly energydependent. A pertinent depth to evaluate the likelihood ofclinical effect is the epidermal/dermal border, defined in thisstudy as 60 �m depth in water. At this depth, there is sig-nificant edge dose enhancement for beams above 100 kVp.A 150 kVp beam has approximately double edge dose en-hancement increasing to three times edge dose enhancementfor 300 kVp. Plastic film such as gladwrap™ shieldingwrapped around the lead cutout, or placed under the leadcutout, is a simple way to remove the electron contamina-tion. Alternatively, an appropriate thickness of varnish couldbe painted onto the lead.6 Note that if plastic film covers theopening in the lead cutout for the treatment, the plastic film

TABLE I. Layers of gladwrap™ required to reduce edge dose enhancement inthe 10–60 �m epidermal layer to less than 10%.

Tube voltage�kV� HVL

Layersof gladwrap™

Edge dose: Central dosebetween 10 and 60 �m

300 3.8 mm Cu 10 1.10250 2.5 mm Cu 9 1.07180 0.99 mm Cu 5 1.06150 0.60 mm Cu 4 1.09100 6.53 mm Al 2 1.0350 0.79 mm Al 0 1.00

should also be used in dosimetry.

Medical Physics, Vol. 37, No. 8, August 2010

ACKNOWLEDGMENTS

The authors thank Marianne Rinks from the NorthernSydney Cancer Centre for the provision of a serrated leadcutout.

a�Electronic mail: jessica.lye@arpansa.gov.au1S. C. Klevenhagen, D. D’Souza, and I. Bonnefoux, “Complications inlow energy x-ray dosimetry caused by electron contamination,” Phys.Med. Biol. 36, 1111–1116 �1991�.

2P. Andreo, D. T., Burns, K. Hohlfeld, M. S. Huq, T. Kanai, F. Laitano, V.Smyth, and S. Vynckier, “Absorbed dose determination in external beamradiotherapy: An international code of practice for dosimetry based onstandards of absorbed dose to water,” IAEA Technical Report Series No.398, ed. V.11b, 2004.

3C.-M. Ma, C. W. Coffey, L. A. DeWerd, C. Liu, R. Nath, S. M. Seltzer,and J. P. Seuntjens, “AAPM protocol for 40–300 kV x-ray beam dosim-etry in radiotherapy and radiobiology,” Med. Phys. 28, 868–893 �2001�.

4A. S. Shiu, S. S. Tung, R. J. Gastorf, K. R. Hogstrom, W. H. Morrison,and L. J. Peters, “Dosimetric evaluation of lead and tungsten eye shieldsin electron beam treatment,” Int. J. Radiat. Oncol., Biol., Phys. 35, 599–604 �1996�.

5C. H. M. Lee and K. K. D. Chan, “Electron contamination from the leadcutout used in kilovoltage radiotherapy,” Phys. Med. Biol. 45, 1–8�2000�.

6D. W. Thomas and J. A. Clark, “Reducing the skin dose from secondaryelectrons in kilovoltage radiotherapy: A pliable coating for custom leadcut-outs,” J. Radiotherapy in Practice 8, 233–237 �2009�.

7B. Walters, I. Kawrakow, and D. W. O. Rogers, “DOSXYZnrc usersmanual,” NRCC Report No. PIRS-794, rev. B, 2007.

8D. M. Tucker, G. T. Barnes, and D. P. Charkraborty, “Semiempiricalmodel for generating tungsten target x-ray spectra,” Med. Phys. 18, 211–218 �1991�.

9J. E. Lye, D. J. Butler, and D. W. Webb, “Monte Carlo correction factorsfor the ARPANSA kilovoltage free-air chambers and the effect of movingthe limiting aperture,” Metrologia 47, 11–20 �2010�.

10B. Hernández-Machin, L. Borrego, M. Gil-Garcia, and B. H. Hernandez,“Office-based radiation therapy for cutaneous carcinoma: Evaluation of710 treatments,” Int. J. Dermatol. 46, 453–459 �2007�.

11J. T. Whitton and J. D. Everall, “The thickness of the epidermis,” Br. J.Dermatol. 89, 467–476 �1973�.

12T. Gambichler, R. Matip, G. Moussa, P. Altmeyer, and K. Hoffman, “Invivo data of epidermal thickness evaluated by optical coherence tomog-raphy: Effects of age, gender, skin type, and anatomic site,” J. Dermatol.Sci. 44, 145–152 �2006�.

13C. K. Bomford and I. H. Kunkler, Walter and Miller’s Textbook of Radio-therapy: Radiation Physics Therapy and Oncology, 6th ed. �Churchill

Livingstone, Edinburgh, NY, 2003�.