- Email: [email protected]
The effect on dose of kilovoltage x-rays backscattered from lead
In, J Radmlron Oncolo,~~’Bwl. .“hw Vol Pnnted an the U.S.A. All nghts reserved.
0360-3016191 $5.00 t .OO CopyrIght 6 1992 Pergamon Press Ltd.
24. 0~ I1 I - I75
0 Technical Innovations and Notes THE EFFECT ON DOSE OF KILOVOLTAGE BACKSCATI’ERED FROM LEAD M.
SAIFUL HUQ, PH.D., N. VENKATARAMANAN,
Department
of Therapeutic
PH.D.
X-RAYS
AND JEROME
A. MELI?
PH.D.
Radiology, Yale University School of Medicine, 333 Cedar Street, New Haven, CT
Dose enhancement on the backscatter side of a soft tissue/high Z material interface is known to exist for megavoltage x-ray beams. Caused by an increase in backscattered electron fluence, the enhancement persists for short distances upstream of the interface, equal to the range of these electrons. Since photon interaction cross sections are small, there is little photon backscatter at these energies. Consequently, beyond the range of the backscattered electrons, the dose upstream is unaffected by the presence of the interface. A similar dose enhancement has been reported for kilovoltage beams. In this case, due to the very low energy of the backscattered electrons, the enhancement persists for a very short upstream distance. Since photon interaction cross sections at keV energies are relatively large, there is also a substantial backscattered photon fluence. This experimental work investigates the effect of these photons on dose at distances upstream from a water/lead interface beyond the range of the backscattered electrons. Measurements of ionization charge, as a function of interface distance and field size for 60,100, and 250 kV beams, were made with a parallel plate chamber at a fixed depth. A significant underdose was found upstream of the interface compared to a homogeneous water medium. For example, with the 100 kV beam and a 15 X 15 cm2 field the measured underdose is 23% at 3 mm and 14% at 1.5 cm upstream of the interface. The effect decreases with field size. In fact, for a 2 X 2 cm* field the upstream dose in unaffected by the interface. Detailed results for this and the other two beams are presented along with backscatter factor measurements for lead. An explanation for the observed underdose is also presented. Orthovoltage,
Backscatter from high Z.
INTRODUCTION
As discussed by Das and Khan ( 1), the backscatter dose is attributable to photons incident on and backscattered from the high Z material; secondary electrons set in motion in the soft tissue that are incident on and backscattered from the high Z material; bremsstrahlung radiation produced within the high Z material; fluorescent radiation resulting from photoelectric interactions within the high Z material; electrons set in motion and backscattered within the high Z material. It is argued, based on the relatively small photon interaction cross sections, that the dose enhancement is caused primarily by the backscatter of the soft tissue secondary electrons. The short upstream range of the dose enhancement supports the conjecture that the backscattered radiation is primarily electrons. An example of a clinical manifestation of this enhanced dose (1, 3, 4) is the frequently observed intense reaction of mucosa adjacent to gold-filled teeth when irradiating the oral cavity. Prevention of this excess reaction can be accomplished by interposing a low Z material, such as wax,
Dose to tissue near an interface formed by materials of different density and atomic number has been the subject of much investigation for megavoltage electron and x-ray beams and for orthovoltage beams. Two recent articles (1, 2) provide comprehensive bibliographies on the subject. In general it is found that there is an enhancement of dose to soft tissue on the backscatter (upstream) side of a soft tissue/high Z material interface. The most systematic study (1) of this effect was done for the megavoltage range from Co-60 to 24 MV and shows that the maximum dose enhancement (immediately upstream of the interface) increases with atomic number of the heterogeneity from about 8% for bone to 70% for lead and is essentially independent of field size and energy. The dose enhancement decreases rapidly with distance upstream from the interface extending only a few mm for Co-60 to about 1 cm for the 24 MV beam.
Accepted for publication
Reprint requests to: Dr. M. S. Huq, Dept. of Radiation Oncology and Nuclear Medicine, Thomas Jefferson University, 111 So. 1 lth St., Philadelphia, PA 19107. 171
3 December 199 1.
172
I. J.
Radiation
Oncology
0 Biology
0 Physics
of 2 to 3 mm thickness between the mucosa and high Z material. A similar dose enhancement to tissue immediately adjacent to and upstream from higher Z materials has been reported for kilovoltage x-ray beams (4. 5, 6). This, too, is produced by low energy electrons and is also easily eliminated by the use of low V absorbing materials. However, the larger interaction cross sections of keV photons compared to MeV photons results in a considerably greater x-ray component to backscattered radiation in kV beams compared to MV beams. For kV beams, backscatter from a soft tissue/high Z material interface will not only consist of scattered secondary electrons, but also Compton scattered photons and electrons and photoelectrons and fluorescent x-rays resulting from interactions of photons with the high Z material. In this work, the effect on dose to water of the photon component of backscattered radiation from lead is measured for 60, 100, and 250 kV beams as a function of field size and distance from the interface beyond the range of the backscattered electrons. METHODS
AND
MATERIALS
The effect on dose of photons backscattered from lead was measured for 60, 100, and 250 kV beams*, each with a total of 2 mm aluminum filtration and half-value thicknesses of 1.9 mm Al, 3.5 mm Al, and 0.5 mm Cu. respectively. Figure 1 depicts the experimental arrangement. A parallel plate ionization chamber+ was fitted with a I mm thick acrylic cap and placed in a water tank located 60 cm (SSD) from the x-ray source. The 0.25 cc collection volume of this chamber is built into an acrylic cylinder of 3 cm diameter and 3.2 cm height. With the back of the chamber against the inside wall of the tank, the window faced downstream and was located at a depth of 4 cm. Lead of 1.6 mm thickness was suspended from a rail that moved across the top of the tank adjacent to a scale set to indicate the distance of the lead from the chamber window. Ionization charge was collected for interface distances ranging from 3 mm to 20 cm from the fixed chamber window location. Two different cross sections of lead were used, one of 20 X 20 cm* and the other of 3 X 3 cm’. For the larger lead, data was collected for field sizes of 2 X 2, 5 X 5, 10 X 10, and 15 X 15 cm* defined on the phantom surface, all smaller than the lead cross section. For the smaller lead, data was collected for surface field sizes of 5 X 5, 10 X 10 and 15 X 15 cm’, all larger than the lead cross section. To circumvent the problems associated with machine instability and the need for timer error corrections, the
* Siemens Stabilipan. + Memorial Pipe Chamber, NY.
Volume
74, Number
Fig. 1. The experimental setup is depicted here. A thin window parallel plate chamber with a I mm thick acrylic cap is shown in a water tank which is 60 cm from the x-ray target. All data were collected with the chamber window at a fixed 4 cm depth and facing downstream. A 1.6 mm thick piece of lead centered on the beam was moved along the top of the tank. Ionization charge was collected as a function of the distance between the chamber and the water/lead interface for several field sizes. To achieve better reproducibility. the collected charge was referenced to a monitor chamber instead of the x-ray unit timer.
ionization charge was referenced to that of a parallel plate chamber” mounted in a 1.O cm thick polystyrene slab and attached to the head of the x-ray unit. Data presented are the average of at least two readings per interface position which, when referenced to the monitor chamber, were reproducible to better than 0.5%. Ionization charge is taken to be proportional to dose and normalized to that for the largest interface distance (13 cm for the 60 kV beam and 20 cm for the other beams), which is equivalent to a homogeneous water phantom.
RESULTS
Hospital,
New York,
AND DISCUSSION
The dose correction factor is defined to be the ionization charge collected for each interface distance normalized to that for a homogeneous water medium (interface distance of 13 cm for the 60 kV beam and 20 cm for the other beams). Figure 2 gives the correction factor versus distance from the interface as a function of field size for the 60, 100, and 250 kV beams respectively, all with the 20 X 20 cm* lead. Compared to a homogeneous medium, there is a decrease in dose upstream of the interface. The dose correction factor is strongly dependent on field size, with a large dose reduction near the interface for large fields decreasing to virtually no correction for very small fields. The greatest effect was measured for the 100 kV beam. At this energy the dose correction factor 3 mm from the interface is 0.78, 0.81, 0.87, and 0.99 for a 15 X 15, 10 X 10, 5 X 5, and 2 X 2 cm2 field, respectively. The perturbing influence of the lead diminishes with increasing distance from the interface, but not very rapidly. Even 3 cm from the interface, the dose correction factor is about 0.95 for 10 X 10 cm2 field. Similar, but somewhat less
* Holt Chamber, Memorial
1. 1992
Nuclear Assoc. Inc., NY.
Kilovoltage x-rays backscattered from lead 0
/ 5
I
/
I
10
15
20
Distance From Interface (cm)
Fig. 2. The dose correction factor for the 20 X 20 cm2 cross section lead in the 60, 100 and 250 kV beams versus the distance from the water/lead interface. Indicated field sizes are defined on the phantom surface. The dose to a point on the central axis upstream of the interface is that for a homogeneous water medium multiplied by the dose correction factor. The solid lines are a visual fit to the data points.
pronounced, effects were observed for the 60 kV and 250 kV beams. The explanation for the dependence of the dose correction factor on field size lies in the short range of the backscattered photons in lead compared to water. The principal source of backscattered photons in water is the Compton interaction. With lead, backscattered photons originate from Compton interactions and from characteristic radiation following photoelectric absorption of the primary beam. For example, with both lead and water, a 250 keV primary photon produces a backscattered photon spectrum with maximum energy of 125 keV. Photoelectric absorption in lead gives rise to characteristic x-rays of about 10 and 74 keV, which lie within the backscattered Compton spectrum. Thus, lead and water give rise to backscattered photons of similar energies. However, these photons have a very short range in lead. The mean free path of a 125 keV photon is only about 0.3 mm in lead compared to about 63 mm in water. Therefore, at short distances from the
4 PTW, Inc. Freiburg, Germany.
173
interface only a small cross sectional area of lead contributes to the backscattered photon fluence. This area is es-
60 kV
0.8:
M. S. HUQ et al.
sentially independent of field size because photons or& inating from beyond it, as with larger fields, must traverse a longer oblique distance through the lead and are attenuated before reaching the measurement point. So backscatter, and therefore dose, near the interface increases much more slowly with field size than in a homogeneous water medium in which backscattered photons have considerably longer mean free paths. This gives rise to the observed field size dependence of the dose correction factor. As the interface is moved further away from the measurement point, the small effective backscattering area of the lead is replaced by the larger one of water resulting in an increase in backscatter and a smaller reduction in dose. If this explanation is valid, the backscatter factor for lead should vary slowly, if at all, with field size. A 0.3 cc waterproofed thimble chambers was used to measure the backscatter factor of lead and water for the 100 kV beam. The measurements were taken with the chamber in air and with 1 cm lead and 25 cm of water as backscatter material using the standard technique (7) of the chamber axis perpendicular to the beam axis and half-embedded into the backscatter material. The chamber’s lucite wall is thick enough (0.75 mm) to attenuate excess photoelectrons generated in the lead. Ionization charge was taken as the average collected for 1 min and 0.5 min time intervals for field sizes ranging from 3 X 3 to 15 X 15 cm2 defined at the chamber location of 60 cm from the x-ray target. Backscatter factor (BSF) is defined as the ratio of charge collected with the backscatter material to that charge collected in air. As shown in Table 1, the BSF for lead is about 1.07 and independent of the field sizes used, while that for water displays the characteristic increase with field size. Backscatter factors measured for water in this work are in excellent agreement with those commonly used (8). In principle, the dose correction factor for tissue near the lead equals the ratio of the BSF for lead to that for water. This ratio for the 15 X 15 cm2 field is 0.80, which is in good agreement with the 0.77 dose correction factor shown in Figure 3 at 3 mm distance from the lead. Thus, the reduced dose to tissue on the backscatter side of the lead and its dependence on field size is indeed caused by the fact that the BSF for lead is small and independent of field size. Data presented up to this point have been for the lead cross section greater than the radiation field area. In some clinical situations a lead shield of cross section smaller than the field size might be used, for example, to protect teeth while treating a lesion of the cheek. Also, regions of high Z material are frequently present in the oral cavity in teeth containing fillings or fitted with crowns. In these cases, the high Z material partially blocks backscatter ra-
174
I. .I. Radiation
Oncology
0
Biology0 Physics
Table 1. Backscatter factor of a 100 kV (3.5 mm Al HVL) beam for lead and water Field size 3x3 6X6 10 x 10 15 x 15
Lead
Water
1.06
1.10
1.08 1.08
1.24
1.06
1.28 1.32
diation coming from surrounding lower Z tissue. Figure 3 shows the dose correction factor for the 100 and 250 kVp beams as a function of upstream distance from the water/lead interface formed with the 3 X 3 cm* cross section piece of lead. Here the field sizes are larger than the lead cross section. As before, tissue upstream of the interface receives less dose than in a homogeneous water medium. When the field size is not much larger than the lead cross section, the dose correction factor is essentially the same as obtained with the larger lead. As field size increases, more backscatter from water surrounding the small lead reaches the measurement point and the dose reduction is less than obtained with the larger lead. For example, 3 mm from the interface in the 250 kV beam the dose correction factor changes from 0.77 with the large lead to 0.86 with the 3 X 3 cm* lead for the 1.5 X 15 cm2 field, but is unchanged for the 5 X 5 cm* field.
Volume
24, Number
I, I992
factor approaches unity with increasing distance from the interface, it does so rather slowly so that there is still a significant underdose 2 and even 3 cm upstream of the lead. In addition, the underdose diminishes with decreasing field size, such that, for a 2 X 2 cm* field the dose correction factor is essentially 1.O for all energies and distances from the interface. The explanation offered for the underdose of tissue upstream from the lead and the field size dependence of the effect is that because of the short mean free path of keV photons in lead, the backscattered photon fluence originates from only a small area of the lead, even for large field sizes. This is supported by measurements showing that the backscatter factor for lead in a 100 kV beam is independent of field size. A high Z material smaller than the field size partially blocks photons backscattered from the surrounding soft tissue. again resulting in an underdose upstream of the interface. The dose correction factor in this case approaches unity with increasing distance from the interface more rapidly than for the larger cross section lead. If lead is used to shield organs within a kilovoltage radiation field (e.g., to shield teeth when irradiating lesion of the cheek), the calculation of dose to soft tissue upstream of the interface beyond the range of backscattered electrons should incorporate the dose correction factor given in this work to avoid a substantial underdosing of that tissue. Even if no shield is used, the presence of high Z materials, such as in restored teeth, could require the use of similar correction factors and should be considered.
CONCLUSION
When kilovoltage x-rays are incident on a soft tissue/ high Z material interface, there is an increase in the backscattered fluence of both electrons and photons. As reported elsewhere, the electron component produces a dose enhancement to soft tissue adjacent to the interface. This enhancement extends a very short upstream distance equal to the range of the low energy electrons. Adverse reactions from this enhancement can be avoided by interposing a thin low Z material between the lead and the upstream tissue. In this work, the effect on dose of only the photon component of the backscattered radiation from a water/lead interface is presented for 60, 100, and 250 kV beams. Two different cross sections of lead were used (20 X 20 cm* and 3 X 3 cm2), each of 1.6 mm thickness. The measurements show a significant underdose on the backscatter side of the lead well beyond the range of backscattered electrons. The underdose is described as a dose correction factor which is the dose to tissue with the high Z material present to that for a homogeneous water medium. The largest underdose of about 23% was observed for the 100 kV 15 X 15 cm2 beam 3 mm upstream of the interface formed using the larger lead. Although the dose correction
I
O*cb
I
I
I
I
5
10
15
20
Distance FromIntetfaco(cm) Fig. 3. The dose correction factor for the 3 X 3 cm’ cross section lead in the 100 and 250 kV beams versus the distance from the water/lead interface.
Kilovoltage x-rays backscattered from lead 0 M. S. HUQ et al.
175
REFERENCES Das, I. J.; Kahn, F. M. Backscatter dose perturbation at high atomic number interfaces in megavoltage photon beams. Med. Phys. 16: 367-375; 1989. Farahani, M.; Eichmiller, F. C.; McLaughlin, W. L. Measurement of absorbed doses near metal and dental material interfaces irradiated by x- and gamma-ray therapy beams. Phys. Med. Biol. 35: 369-385; 1990. Thambi, V.; Mm-thy, A. K.; Alder, G.; Kartha, P. K. Dose perturbation resulting from gold fillings in patients with head and neck cancers. Int. J. Radiat. Oncol. Biol. Phys. 5: 58 l582; 1979. 4. Gibbs, F. A.; Palos, B.; Goffinet, D. R. The metal/tissue
interface effect in irradiation of the oral cavity. Radio]. 119: 705-707; 1976. Wingate, C. L.; Gross, W.; Failla, G. Experimental determination of absorbed dose from x-rays near the inface of soft tissue and other materials. Radiol. 79: 984- 1000; 1962. Hood, S. L.; Norris, G. Dosimetry of human cell cultures irradiated at the interface in plastic and glass dishes. Radiat. Res. 14: 705-712; 1961. Johns, H. E.; Hunt, J. W.; Fedoruk, S. 0. Surface backscatter in the 100 kV to 400 kV range. Brit. J. Radiol. 27: 443448; 1954. Br. J. Radiol. (Suppl. 17). Central axis depth dose data for use in radiotherapy, 1983.
0360-3016191 $5.00 t .OO CopyrIght 6 1992 Pergamon Press Ltd.
24. 0~ I1 I - I75
0 Technical Innovations and Notes THE EFFECT ON DOSE OF KILOVOLTAGE BACKSCATI’ERED FROM LEAD M.
SAIFUL HUQ, PH.D., N. VENKATARAMANAN,
Department
of Therapeutic
PH.D.
X-RAYS
AND JEROME
A. MELI?
PH.D.
Radiology, Yale University School of Medicine, 333 Cedar Street, New Haven, CT
Dose enhancement on the backscatter side of a soft tissue/high Z material interface is known to exist for megavoltage x-ray beams. Caused by an increase in backscattered electron fluence, the enhancement persists for short distances upstream of the interface, equal to the range of these electrons. Since photon interaction cross sections are small, there is little photon backscatter at these energies. Consequently, beyond the range of the backscattered electrons, the dose upstream is unaffected by the presence of the interface. A similar dose enhancement has been reported for kilovoltage beams. In this case, due to the very low energy of the backscattered electrons, the enhancement persists for a very short upstream distance. Since photon interaction cross sections at keV energies are relatively large, there is also a substantial backscattered photon fluence. This experimental work investigates the effect of these photons on dose at distances upstream from a water/lead interface beyond the range of the backscattered electrons. Measurements of ionization charge, as a function of interface distance and field size for 60,100, and 250 kV beams, were made with a parallel plate chamber at a fixed depth. A significant underdose was found upstream of the interface compared to a homogeneous water medium. For example, with the 100 kV beam and a 15 X 15 cm2 field the measured underdose is 23% at 3 mm and 14% at 1.5 cm upstream of the interface. The effect decreases with field size. In fact, for a 2 X 2 cm* field the upstream dose in unaffected by the interface. Detailed results for this and the other two beams are presented along with backscatter factor measurements for lead. An explanation for the observed underdose is also presented. Orthovoltage,
Backscatter from high Z.
INTRODUCTION
As discussed by Das and Khan ( 1), the backscatter dose is attributable to photons incident on and backscattered from the high Z material; secondary electrons set in motion in the soft tissue that are incident on and backscattered from the high Z material; bremsstrahlung radiation produced within the high Z material; fluorescent radiation resulting from photoelectric interactions within the high Z material; electrons set in motion and backscattered within the high Z material. It is argued, based on the relatively small photon interaction cross sections, that the dose enhancement is caused primarily by the backscatter of the soft tissue secondary electrons. The short upstream range of the dose enhancement supports the conjecture that the backscattered radiation is primarily electrons. An example of a clinical manifestation of this enhanced dose (1, 3, 4) is the frequently observed intense reaction of mucosa adjacent to gold-filled teeth when irradiating the oral cavity. Prevention of this excess reaction can be accomplished by interposing a low Z material, such as wax,
Dose to tissue near an interface formed by materials of different density and atomic number has been the subject of much investigation for megavoltage electron and x-ray beams and for orthovoltage beams. Two recent articles (1, 2) provide comprehensive bibliographies on the subject. In general it is found that there is an enhancement of dose to soft tissue on the backscatter (upstream) side of a soft tissue/high Z material interface. The most systematic study (1) of this effect was done for the megavoltage range from Co-60 to 24 MV and shows that the maximum dose enhancement (immediately upstream of the interface) increases with atomic number of the heterogeneity from about 8% for bone to 70% for lead and is essentially independent of field size and energy. The dose enhancement decreases rapidly with distance upstream from the interface extending only a few mm for Co-60 to about 1 cm for the 24 MV beam.
Accepted for publication
Reprint requests to: Dr. M. S. Huq, Dept. of Radiation Oncology and Nuclear Medicine, Thomas Jefferson University, 111 So. 1 lth St., Philadelphia, PA 19107. 171
3 December 199 1.
172
I. J.
Radiation
Oncology
0 Biology
0 Physics
of 2 to 3 mm thickness between the mucosa and high Z material. A similar dose enhancement to tissue immediately adjacent to and upstream from higher Z materials has been reported for kilovoltage x-ray beams (4. 5, 6). This, too, is produced by low energy electrons and is also easily eliminated by the use of low V absorbing materials. However, the larger interaction cross sections of keV photons compared to MeV photons results in a considerably greater x-ray component to backscattered radiation in kV beams compared to MV beams. For kV beams, backscatter from a soft tissue/high Z material interface will not only consist of scattered secondary electrons, but also Compton scattered photons and electrons and photoelectrons and fluorescent x-rays resulting from interactions of photons with the high Z material. In this work, the effect on dose to water of the photon component of backscattered radiation from lead is measured for 60, 100, and 250 kV beams as a function of field size and distance from the interface beyond the range of the backscattered electrons. METHODS
AND
MATERIALS
The effect on dose of photons backscattered from lead was measured for 60, 100, and 250 kV beams*, each with a total of 2 mm aluminum filtration and half-value thicknesses of 1.9 mm Al, 3.5 mm Al, and 0.5 mm Cu. respectively. Figure 1 depicts the experimental arrangement. A parallel plate ionization chamber+ was fitted with a I mm thick acrylic cap and placed in a water tank located 60 cm (SSD) from the x-ray source. The 0.25 cc collection volume of this chamber is built into an acrylic cylinder of 3 cm diameter and 3.2 cm height. With the back of the chamber against the inside wall of the tank, the window faced downstream and was located at a depth of 4 cm. Lead of 1.6 mm thickness was suspended from a rail that moved across the top of the tank adjacent to a scale set to indicate the distance of the lead from the chamber window. Ionization charge was collected for interface distances ranging from 3 mm to 20 cm from the fixed chamber window location. Two different cross sections of lead were used, one of 20 X 20 cm* and the other of 3 X 3 cm’. For the larger lead, data was collected for field sizes of 2 X 2, 5 X 5, 10 X 10, and 15 X 15 cm* defined on the phantom surface, all smaller than the lead cross section. For the smaller lead, data was collected for surface field sizes of 5 X 5, 10 X 10 and 15 X 15 cm’, all larger than the lead cross section. To circumvent the problems associated with machine instability and the need for timer error corrections, the
* Siemens Stabilipan. + Memorial Pipe Chamber, NY.
Volume
74, Number
Fig. 1. The experimental setup is depicted here. A thin window parallel plate chamber with a I mm thick acrylic cap is shown in a water tank which is 60 cm from the x-ray target. All data were collected with the chamber window at a fixed 4 cm depth and facing downstream. A 1.6 mm thick piece of lead centered on the beam was moved along the top of the tank. Ionization charge was collected as a function of the distance between the chamber and the water/lead interface for several field sizes. To achieve better reproducibility. the collected charge was referenced to a monitor chamber instead of the x-ray unit timer.
ionization charge was referenced to that of a parallel plate chamber” mounted in a 1.O cm thick polystyrene slab and attached to the head of the x-ray unit. Data presented are the average of at least two readings per interface position which, when referenced to the monitor chamber, were reproducible to better than 0.5%. Ionization charge is taken to be proportional to dose and normalized to that for the largest interface distance (13 cm for the 60 kV beam and 20 cm for the other beams), which is equivalent to a homogeneous water phantom.
RESULTS
Hospital,
New York,
AND DISCUSSION
The dose correction factor is defined to be the ionization charge collected for each interface distance normalized to that for a homogeneous water medium (interface distance of 13 cm for the 60 kV beam and 20 cm for the other beams). Figure 2 gives the correction factor versus distance from the interface as a function of field size for the 60, 100, and 250 kV beams respectively, all with the 20 X 20 cm* lead. Compared to a homogeneous medium, there is a decrease in dose upstream of the interface. The dose correction factor is strongly dependent on field size, with a large dose reduction near the interface for large fields decreasing to virtually no correction for very small fields. The greatest effect was measured for the 100 kV beam. At this energy the dose correction factor 3 mm from the interface is 0.78, 0.81, 0.87, and 0.99 for a 15 X 15, 10 X 10, 5 X 5, and 2 X 2 cm2 field, respectively. The perturbing influence of the lead diminishes with increasing distance from the interface, but not very rapidly. Even 3 cm from the interface, the dose correction factor is about 0.95 for 10 X 10 cm2 field. Similar, but somewhat less
* Holt Chamber, Memorial
1. 1992
Nuclear Assoc. Inc., NY.
Kilovoltage x-rays backscattered from lead 0
/ 5
I
/
I
10
15
20
Distance From Interface (cm)
Fig. 2. The dose correction factor for the 20 X 20 cm2 cross section lead in the 60, 100 and 250 kV beams versus the distance from the water/lead interface. Indicated field sizes are defined on the phantom surface. The dose to a point on the central axis upstream of the interface is that for a homogeneous water medium multiplied by the dose correction factor. The solid lines are a visual fit to the data points.
pronounced, effects were observed for the 60 kV and 250 kV beams. The explanation for the dependence of the dose correction factor on field size lies in the short range of the backscattered photons in lead compared to water. The principal source of backscattered photons in water is the Compton interaction. With lead, backscattered photons originate from Compton interactions and from characteristic radiation following photoelectric absorption of the primary beam. For example, with both lead and water, a 250 keV primary photon produces a backscattered photon spectrum with maximum energy of 125 keV. Photoelectric absorption in lead gives rise to characteristic x-rays of about 10 and 74 keV, which lie within the backscattered Compton spectrum. Thus, lead and water give rise to backscattered photons of similar energies. However, these photons have a very short range in lead. The mean free path of a 125 keV photon is only about 0.3 mm in lead compared to about 63 mm in water. Therefore, at short distances from the
4 PTW, Inc. Freiburg, Germany.
173
interface only a small cross sectional area of lead contributes to the backscattered photon fluence. This area is es-
60 kV
0.8:
M. S. HUQ et al.
sentially independent of field size because photons or& inating from beyond it, as with larger fields, must traverse a longer oblique distance through the lead and are attenuated before reaching the measurement point. So backscatter, and therefore dose, near the interface increases much more slowly with field size than in a homogeneous water medium in which backscattered photons have considerably longer mean free paths. This gives rise to the observed field size dependence of the dose correction factor. As the interface is moved further away from the measurement point, the small effective backscattering area of the lead is replaced by the larger one of water resulting in an increase in backscatter and a smaller reduction in dose. If this explanation is valid, the backscatter factor for lead should vary slowly, if at all, with field size. A 0.3 cc waterproofed thimble chambers was used to measure the backscatter factor of lead and water for the 100 kV beam. The measurements were taken with the chamber in air and with 1 cm lead and 25 cm of water as backscatter material using the standard technique (7) of the chamber axis perpendicular to the beam axis and half-embedded into the backscatter material. The chamber’s lucite wall is thick enough (0.75 mm) to attenuate excess photoelectrons generated in the lead. Ionization charge was taken as the average collected for 1 min and 0.5 min time intervals for field sizes ranging from 3 X 3 to 15 X 15 cm2 defined at the chamber location of 60 cm from the x-ray target. Backscatter factor (BSF) is defined as the ratio of charge collected with the backscatter material to that charge collected in air. As shown in Table 1, the BSF for lead is about 1.07 and independent of the field sizes used, while that for water displays the characteristic increase with field size. Backscatter factors measured for water in this work are in excellent agreement with those commonly used (8). In principle, the dose correction factor for tissue near the lead equals the ratio of the BSF for lead to that for water. This ratio for the 15 X 15 cm2 field is 0.80, which is in good agreement with the 0.77 dose correction factor shown in Figure 3 at 3 mm distance from the lead. Thus, the reduced dose to tissue on the backscatter side of the lead and its dependence on field size is indeed caused by the fact that the BSF for lead is small and independent of field size. Data presented up to this point have been for the lead cross section greater than the radiation field area. In some clinical situations a lead shield of cross section smaller than the field size might be used, for example, to protect teeth while treating a lesion of the cheek. Also, regions of high Z material are frequently present in the oral cavity in teeth containing fillings or fitted with crowns. In these cases, the high Z material partially blocks backscatter ra-
174
I. .I. Radiation
Oncology
0
Biology0 Physics
Table 1. Backscatter factor of a 100 kV (3.5 mm Al HVL) beam for lead and water Field size 3x3 6X6 10 x 10 15 x 15
Lead
Water
1.06
1.10
1.08 1.08
1.24
1.06
1.28 1.32
diation coming from surrounding lower Z tissue. Figure 3 shows the dose correction factor for the 100 and 250 kVp beams as a function of upstream distance from the water/lead interface formed with the 3 X 3 cm* cross section piece of lead. Here the field sizes are larger than the lead cross section. As before, tissue upstream of the interface receives less dose than in a homogeneous water medium. When the field size is not much larger than the lead cross section, the dose correction factor is essentially the same as obtained with the larger lead. As field size increases, more backscatter from water surrounding the small lead reaches the measurement point and the dose reduction is less than obtained with the larger lead. For example, 3 mm from the interface in the 250 kV beam the dose correction factor changes from 0.77 with the large lead to 0.86 with the 3 X 3 cm* lead for the 1.5 X 15 cm2 field, but is unchanged for the 5 X 5 cm* field.
Volume
24, Number
I, I992
factor approaches unity with increasing distance from the interface, it does so rather slowly so that there is still a significant underdose 2 and even 3 cm upstream of the lead. In addition, the underdose diminishes with decreasing field size, such that, for a 2 X 2 cm* field the dose correction factor is essentially 1.O for all energies and distances from the interface. The explanation offered for the underdose of tissue upstream from the lead and the field size dependence of the effect is that because of the short mean free path of keV photons in lead, the backscattered photon fluence originates from only a small area of the lead, even for large field sizes. This is supported by measurements showing that the backscatter factor for lead in a 100 kV beam is independent of field size. A high Z material smaller than the field size partially blocks photons backscattered from the surrounding soft tissue. again resulting in an underdose upstream of the interface. The dose correction factor in this case approaches unity with increasing distance from the interface more rapidly than for the larger cross section lead. If lead is used to shield organs within a kilovoltage radiation field (e.g., to shield teeth when irradiating lesion of the cheek), the calculation of dose to soft tissue upstream of the interface beyond the range of backscattered electrons should incorporate the dose correction factor given in this work to avoid a substantial underdosing of that tissue. Even if no shield is used, the presence of high Z materials, such as in restored teeth, could require the use of similar correction factors and should be considered.
CONCLUSION
When kilovoltage x-rays are incident on a soft tissue/ high Z material interface, there is an increase in the backscattered fluence of both electrons and photons. As reported elsewhere, the electron component produces a dose enhancement to soft tissue adjacent to the interface. This enhancement extends a very short upstream distance equal to the range of the low energy electrons. Adverse reactions from this enhancement can be avoided by interposing a thin low Z material between the lead and the upstream tissue. In this work, the effect on dose of only the photon component of the backscattered radiation from a water/lead interface is presented for 60, 100, and 250 kV beams. Two different cross sections of lead were used (20 X 20 cm* and 3 X 3 cm2), each of 1.6 mm thickness. The measurements show a significant underdose on the backscatter side of the lead well beyond the range of backscattered electrons. The underdose is described as a dose correction factor which is the dose to tissue with the high Z material present to that for a homogeneous water medium. The largest underdose of about 23% was observed for the 100 kV 15 X 15 cm2 beam 3 mm upstream of the interface formed using the larger lead. Although the dose correction
I
O*cb
I
I
I
I
5
10
15
20
Distance FromIntetfaco(cm) Fig. 3. The dose correction factor for the 3 X 3 cm’ cross section lead in the 100 and 250 kV beams versus the distance from the water/lead interface.
Kilovoltage x-rays backscattered from lead 0 M. S. HUQ et al.
175
REFERENCES Das, I. J.; Kahn, F. M. Backscatter dose perturbation at high atomic number interfaces in megavoltage photon beams. Med. Phys. 16: 367-375; 1989. Farahani, M.; Eichmiller, F. C.; McLaughlin, W. L. Measurement of absorbed doses near metal and dental material interfaces irradiated by x- and gamma-ray therapy beams. Phys. Med. Biol. 35: 369-385; 1990. Thambi, V.; Mm-thy, A. K.; Alder, G.; Kartha, P. K. Dose perturbation resulting from gold fillings in patients with head and neck cancers. Int. J. Radiat. Oncol. Biol. Phys. 5: 58 l582; 1979. 4. Gibbs, F. A.; Palos, B.; Goffinet, D. R. The metal/tissue
interface effect in irradiation of the oral cavity. Radio]. 119: 705-707; 1976. Wingate, C. L.; Gross, W.; Failla, G. Experimental determination of absorbed dose from x-rays near the inface of soft tissue and other materials. Radiol. 79: 984- 1000; 1962. Hood, S. L.; Norris, G. Dosimetry of human cell cultures irradiated at the interface in plastic and glass dishes. Radiat. Res. 14: 705-712; 1961. Johns, H. E.; Hunt, J. W.; Fedoruk, S. 0. Surface backscatter in the 100 kV to 400 kV range. Brit. J. Radiol. 27: 443448; 1954. Br. J. Radiol. (Suppl. 17). Central axis depth dose data for use in radiotherapy, 1983.