Perez & Brady's Principles and Practice of Radiation Oncology (Perez and Bradys Principles and Practice of Radiation Oncology), 6 Ed.

Chapter 7. Photon External-Beam Dosimetry and Treatment Planning

James A. Purdy and Eric E. Klein

The radiation oncologist, when planning the treatment of a patient with cancer, is faced with the problem of prescribing a treatment regimen with a radiation dose that is large enough potentially to cure or control the disease, but does not cause serious normal tissue complications. This task is a difficult one because tumor control and normal tissue effect responses for most disease sites are typically steep functions of radiation dose; that is, a small change in the dose delivered (±5%) can result in a dramatic change in the local response of the tissue (±20%).^1,2,3 Moreover, the prescribed curative doses are often, by necessity, very close to the doses tolerated by the normal tissues. Thus, for optimum treatment, the radiation dose must be planned and delivered with a high degree of precision.

One can readily compute the dose distribution resulting from photons, electrons, protons, or a mixture of these radiation beams impinging on a regularly shaped, flat-surface, homogeneous unit-density phantom. However, the patient presents a much more complicated situation because of irregularly shaped topography and having tissues of varying densities and atomic composition (called heterogeneities). In addition, beam modifiers, such as wedges, compensating filters, or bolus, are sometimes inserted into the radiation beam, further complicating the calculation of the absorbed dose.

In this chapter, several aspects of photon external-beam treatment planning and dosimetry are reviewed, including methods used for dose/monitor unit calculations, correction for the effects of the patient’s irregular surface and internal heterogeneities on the calculated photon dose distribution, isodose distributions for combined fields, field junctions, field shaping and design of treatment aids, and related clinical dosimetry issues.

DOSE CALCULATION METHODS

For purposes of discussion, it is convenient to characterize photon beam dose calculation methods as either correction based or model based.⁴ In the former method, the dose at a given point is calculated using measured central-axis data, for example, percent depth dose (PDD), tissue-air ratios (TARs), tissue-maximum ratios (TMRs), tissue-phantom ratios (TPRs), and off-axis ratios (OARs). These quantities are measured under reference conditions (i.e., in a homogeneous water phantom with a flat surface normal to the incident radiation beam at a standard distance from the x-ray source). Hence, specific correction factors (CFs) are used in planning the treatment of real patients to account for varying patient surfaces, tissue heterogeneities, irregular field shapes, and any beam modifiers used.

In the model-based algorithms, the dose distribution is computed in a phantom or patient from more of a first-principles approach accounting for lateral transport of radiation, beam energy, geometry, beam modifiers, patient surface topography, and electron density distribution, rather than correcting parameterized dose distributions measured in a water phantom. These models utilize convolution energy deposition kernels that describe the distribution of dose about a single primary photon interaction site, and provide much more accurate results even for complex heterogeneous geometries. Both methods are discussed in the following sections and more details on model-based photon dose calculation algorithms can be found in the references listed.^4,5

Correction-Based Dose Calculation Methods

Using the notation of Khan et al.,⁶ the dose at a point (D_P) at a depth d of overlying tissue on the central ray for an irregularly shaped field is given by:

where S_c denotes the collimator scatter factor, Sp the phantom scatter factor, and TMR the tissue-maximum ratio. TF and WF denote the tray and wedge factors, respectively, and are defined as the ratio of the central ray dose with the tray or wedge filter in place relative to the dose in the open-beam geometry. The collimated field size is denoted by r_c and is usually described as the square field size equivalent to the rectangular collimator opening projected to isocenter. The effective field size is denoted by r_d and is specified to the isocenter distance (SAD). The inverse-square law factor accounts for the difference in distances from the source-to-point of dose calculation relative to the source to calibration point distance (SCD). Note that when isocentric calibration is used, this factor is unity. Also, note that collimator-defined field size is used for lookup of the collimator scatter factor, S_c, whereas effective field size projected to isocenter is used for lookup of the TMR value and the phantom scatter factor, Sp. By separately accounting for the effect of collimator opening on the primary dose component and the influence of cross-sectional area of tissue irradiated, most of the difficulties in accurately calculating a dose in the presence of extensive blocking are overcome. Details on determining the effective field size for an irregularly shaped field, taking into account both the primary and scatter dose components, will be discussed in a later section.

Correction for Varying Patient Topography (Air Gaps)

In the previous equation, it is assumed that the beam is normally incident on a unit-density uniform phantom. The following CF methods can be applied to the equation to account for the nonnormal beam incidence caused by the patient’s varying surface.⁷

FIGURE 7.1. Schematic drawing illustrating tissue-air ratio and effective source-to-skin distance (SSD) methods for the correction of isodose curves under a sloping surface (solid lines for SSD = S′′; dashed lines for SSD = S′′). (From International Commission of Radiation Units and Measurements. Report 24: Determination of absorbed dose in a patient irradiated by beams of × or gamma rays in radiotherapy procedures. Washington, DC: International Commission of Radiation Units and Measurements, 1976.)

Air Gap CF: Ratio of Tissue-Air Ratio Method

In the ratio of TAR or TPR method, the surface (along a ray line) directly above point A (source-to-skin distance [SSD] = S′) is unaltered, so the primary component to the dose distribution at this point is unchanged (and the scatter component is also assumed to be unaltered) (see Fig. 7.1). Thus, the dose at point A can be considered as unaltered by patient shape. However, for point B, where there are considerable variations in the patient’s topography, both the primary and scatter components of the radiation beam are altered. The CF may be determined using two TARs or TPRs as follows:

where h = air gap.

Air Gap CF: Effective Source-Skin Ratio Method

In the effective SSD method, the isodose chart to be used is placed on the patient’s contour representation, positioning the central axis at the distance for which the curve was measured (Fig. 7.1). It then is shifted down along the ray line for the length of the air gap, h, resulting in SSD = S″. The PDD value at point B is read and modified by an inverse-square calculation to account for the effective change in the peak dose. The CF can be expressed as follows:

Correction for Tissue Heterogeneities

The following CF methods can be used to account for the tissue heterogeneities found within the patient.

Tissue Heterogeneities CF: Ratio of Tissue-Air Ratio Method

The ratio of TAR (RTAR) method of correction for inhomogeneities is given by:

where the numerator is the TAR for the equivalent water thickness, d_eff, and the denominator is the TAR for the actual thickness, d, of tissue between the point of calculation and the surface along a ray passing through the point. S_d is the dimension of the beam cross-section at the depth of calculation. The RTAR method accounts for the field size and depth of calculation. It does not account for the position of the point of calculation with respect to the heterogeneity. It also does not take into account the shape of the inhomogeneity; instead, it assumes that it extends the full width of the beam and has a constant thickness (i.e., referred to as slab geometry).

Tissue Heterogeneities CF: Power Law TAR Method

The power law TAR method was proposed by Batho⁸ and generalized by Young and Gaylord.⁹ This method, sometimes called the Batho method, attempts to account for the nature of the inhomogeneity and its position relative to the point of calculation. However, it does not account for the extent or shape of the inhomogeneity. The correction factor for the point P is given by:

where d₁ and d₂ refer to the distances from point P to the near and far side of the non–water-equivalent material, respectively; S_d is the beam dimension at the depth of P; and ρ₂ is the relative electron density of the inhomogeneity with respect to water.

Sontag and Cunningham¹⁰ derived a more general form of this correction factor, which can be applied to a case in which the effective atomic number of the inhomogeneity is different from that of water and the point of interest lies within the inhomogeneity. The correction factor in this situation is given by:

where ρ_a is the density of the material in which point P lies at a depth d below the surface and ρ_b is the density of an overlying material of thickness (d₂ – d₁); (μ_en/ρ)_a and (μ_en/ρ)_b are the mass energy absorption coefficients for the medium a and b.

Model-Based Dose Calculation Methods

Advanced three-dimensional dose calculation algorithms, such as the convolution/superposition algorithm and Monte Carlo, should now be considered the standard of practice.^4,5,11–13 These models provide accurate results even for complex heterogeneous geometries.

Convolution/Superposition Dose Calculation Algorithm

The convolution/superposition dose calculation algorithm is based on the following equation⁴:

where D represents the dose at some point , T_E() represents the total energy released by primary photon interactions per unit mass (or TERMA), and h(E, – ) is the point-spread function (also called dose spread array, differential pencil beam, and energy deposition kernal). The point-spread function represents the fraction of the energy deposited (per unit volume) at point that is subsequently transported to the calculation point, . Hence, the dose at point is computed by integrating over all space the contributions from photons and electrons produced at all other points in the phantom or patient.

Ahnesjö et al.¹⁴ showed that the point-spread function, h(E, – ), changes only slightly as a function of energy, and thus, can be replaced by h( – ) (defined as the average point-spread function weighted by the spectral components of the beam), reducing the basic convolution four-dimensional integral to a three-dimensional integral over all space. Point-spread functions for monoenergetic photons are generally precomputed using Monte Carlo methods.¹⁴ The energy dependence of the TERMA, T_E (), can be expressed by applying the inverse-square law and exponential attenuation to the photon fluence at the surface of the phantom or patient.

The three-dimensional integral is typically evaluated in a two-step process. The first step takes into account the properties of the accelerator (including the finite source size, primary collimator, flattening filter, collimator jaws, multileaf collimators, and any beam-modifying devices used for the treatment, such as wedges, alloy blocks, and compensating filters) to compute the energy fluence at the phantom or patient surface. The second step of the calculation takes into account the inverse-square law and exponential attenuation to this incident fluence to determine the TERMA, T_E(), at each point within the phantom or patient and convolve the result with the point-spread function, h(–).

The convolution equation is strictly valid only for homogeneous media (i.e., h() must be spatially invariant). To account for the effects of tissue heterogeneities, all physical distances in the convolution integral are replaced with radiologic distances, that is, the physical distance multiplied by the average density along the line in question.^11,14 Hence, the convolution/superposition algorithm accounts for the effects of heterogeneities anywhere in the vicinity of the calculation point in three dimensions. In contrast, most correction factor–based dose-calculation techniques require only a simple one-dimensional evaluation of radiologic path length, and can thus account for the effects of only those tissue heterogeneities that lie along a ray connecting the radiation source to the calculation point.

Several investigators have tested the convolution/superposition algorithm against measurements and Monte Carlo–generated data for complex phantom geometries including both homogeneous and heterogeneous phantoms and found that the convolution/superposition model gave accurate results, even in parts of the buildup region and penumbra.^15,16

Monte Carlo Method

Monte Carlo is, in principle, the only method capable of computing the dose distribution accurately for all situations encountered in radiation therapy, including being able to accurately predict the dose near interfaces of materials with very dissimilar atomic number, such as near metal prostheses, or different densities such as tumors in lung tissue.¹⁷ The Monte Carlo method uses the known cross-sections for electron and photon interactions in matter and follows individual photons and the associated electrons set in motion through the entire heterogeneous phantom or patient. By calculating the trajectories and interactions of a very large number of photons and electrons, one can accurately model the dose distribution. Recently, several Monte Carlo codes have been developed for radiotherapy treatment planning,^13,18 many of which have been implemented commercially. The reader is referred to the American Association of Physicists in Medicine (AAPM) Task Group (TG) 105 Report, which summarizes commercial use of Monte Carlo for radiation therapy treatment planning.¹² The reader is also referred to the review article by Siebers et al. for even more details on Monte Carlo calculation for external-beam radiation therapy.¹⁷

Dose Calculation Algorithms and Tissue Heterogeneities

In 2004, the AAPM published Report 85 (Task Group 65) on tissue inhomogeneity corrections for megavoltage photon beams.¹⁹ The task group recommended an accuracy goal for tissue heterogeneity corrections of 2% in order to achieve an overall 3% accuracy in dose delivery. The AAPM report recommended heterogeneity corrections be applied to plans and prescriptions, with the condition that the algorithm used for calculations be reviewed and rigorously tested by the medical physicist. A brief summary of the site-specific recommendations follows. For the head and neck region, a one-dimensional path correction algorithm for point-dose estimations beyond mandible and ear cavities was thought to be reasonable. However, for soft tissue regions and volumes that are adjacent to these heterogeneities, superposition/convolution or Monte Carlo algorithms should be used. For the larynx, specifically, if the target volume was adjacent to the air cavity or severe case of disease in the anterior commissure, then the superposition/convolution or Monte Carlo algorithms should be used. For treatment of lung cancer, for interest points well beyond the lung interface, one-dimensional path corrections were thought to be reasonable. However, accounting for doses at tumor–lung interfaces, the superposition/convolution or Monte Carlo algorithms should be used. Also, the report recommended that photon energies of 12 MV or less should be used for treatment of lung cancer in order to minimize nonequilibrium conditions that exist with higher energies. For breast cancer treatment (particularly if the dose of interest of the target volume is considered to be chest wall), it is recommended that calculations be performed with superposition/convolution or Monte Carlo. However, for simple intact breast planning, one-dimensional algorithms were adequate. For the upper gastrointestinal tract, one-dimensional corrections were adequate. However, one should be leery if barium contrast is used as it can erroneously impact the dose calculation due to its high-Z. In terms of the pelvis and prostate, one-dimensional corrections were quite reasonable except in the presence of high-Z implanted hip prostheses. (Note, the dosimetric considerations for patients with hip prostheses undergoing pelvic irradiation are discussed in a later section). The study by Frank et al.²⁰ provides a clear method for safely transitioning clinical use from one based on planning that assumes a homogeneous unit-density patient to one using a heterogeneous patient model.

Recently the Radiological Physics Center published the results of a study comparing measured results of irradiated lung phantoms having various geometries with dose calculations for similar conditions using commercial treatment planning systems. They found significant differences if algorithms less sophisticated than the superposition/convolution-type algorithms were used.²¹

MONITOR UNIT CALCULATION METHODS

Monitor unit (MU) calculations refer to determining the linac MU setting per field to deliver the prescribed dose taking into account the tumor depth, treatment distance, multileaf collimator setting or secondary blocking configuration, and primary collimator opening. This is accomplished by using the various dosimetric quantities described in the preceding chapter to relate the dose corresponding to an arbitrary set of treatment parameters to the reference calibration geometry where the output of the machine is specified in terms of cGy/MU. The reference source to calibration point distance, field size, and depth of output specification are denoted by the symbols SCD, r_cal, and d_cal, respectively. For a fixed SSD calibration geometry:

Normal incidence and open-beam geometry (i.e., absence of trays or any beam-modifying filters) are specified.

For treatment machines calibrated isocentrically, the point of MU specification is located at distance SAD rather than at distance SAD + d_max as stated previously. For isocentric calibration, SCD = SAD.

The linac is calibrated by adjusting the sensitivity of its internal monitor transmission ion chamber so that 1 MU equals 1 cGy for the reference calibration geometry condition. Several reports providing more details on monitor unit calculations and their verification are listed in the references.^22–24

MU Calculation for Fixed Fields

When the patient is to be treated isocentrically, the point of dose prescription is located at the isocenter regardless of the target depth. Using the notation of Khan et al.,⁶ the MU needed to deliver a prescribed tumor dose to isocenter (TD_iso) for a depth d of overlying tissue on the central ray is given by:

where TF and WF denote the tray and wedge factors, respectively. They are defined as the ratio of the central ray dose with the tray or wedge filter in place relative to the dose in the open-beam geometry. The collimated field size is denoted by r_c and is usually described as the square field size equivalent to the rectangular collimator opening projected to isocenter. The effective field size is denoted by r_d and is always specified to the isocenter distance (SAD). The inverse-square law factor accounts for the difference in distances from the source-to-point of dose prescription relative to the point of MU specification. When isocentric calibration is used, this factor is unity. Note that collimator-defined field size is used for lookup of the collimator scatter factor, S_c, whereas effective field size projected to isocenter is used for lookup of TMR and the phantom scatter factor, S_p. By separately accounting for the effect of collimator opening on the primary dose component and the influence of cross-sectional area of tissue irradiated, most of the difficulties in accurately delivering a dose in the presence of extensive blocking are overcome.

When a fixed distance between the target and entry skin surface (SSD) is used to treat the patient, a dose-calculation formalism based on PDD is used rather than one based on isocentric dose ratios. When a dose TD is to be delivered to depth d, MUs are given by:

The field size (or its equivalent square) on the skin surface at central axis is denoted by r and is used for lookup of both PDD and S_p. The collimated field size r_c at the isocenter must be used for lookup of S_c. When an extended treatment distance is used, the collimated field size at isocenter differs significantly from that at the skin surface of the patient. Note that PDD is a function of SSD, depth, and effective field size. Collimator scatter factors measured at SAD are valid over a wide range of extended treatment distances.⁶

When this dose calculation formalism for highly extended treatment distances such as encountered in administering total-body irradiation is used, care must be taken to verify the validity of inverse-square law at these distances. It is recommended that such setups always be verified by ion chamber measurement at the extended distance. Because of the large scatter contribution to effective primary dose originating from the flattening filter and other components in the treatment head, the virtual source of radiation may be as much as 2 cm proximal to the target of the accelerator.

The TAR system of dose calculation is a widely used alternative to the Khan formalism. It is simply an extension of the familiar TAR and backscatter factor concepts, as used in ⁶⁰Co and orthovoltage dosimetry, to the megavoltage photon energy range. The needed dosimetry parameters are determined from ion chamber measurements (both in-phantom and in-air) like those performed for ⁶⁰Co, but now using a much larger buildup cap (radius thickness = d_max). Thus, the megavoltage peakscatter factor, PSF(r), for an effective field size r is simply the ratio of the two ion chamber readings as shown here.

And the megavoltage beam dose rate (Gy/MU) in free space, , is given by:

where the numerator is the measured d_max dose at distance SSD = SAD + d_max and collimator setting r_c. Implementation of this system requires a table of values for each collimator opening and a table of PSF versus effective field size. Then, dose at d_max per MU for any distance, effective field size, and collimator opening can be calculated easily.

When the patient is to be treated isocentrically, the MU needed to deliver a prescribed isocenter dose (ID) to a depth d on the central axis is given by:

If the treatment is fixed SSD, the MU needed to deliver a prescribed dose (TD) to a depth d on the central ray is given by:

All MU calculation formalisms require some means of estimating the square field size, r, that is equivalent, in terms of scattering characteristics, to an arbitrary rectangular field of width a and length b.Perhaps the most widely used rectangular equivalency principle is the “A/P” rule. It states that a square and a rectangle are equivalent if they have the same area/perimeter ratio; that is:

Another widely used approach to reducing rectangular estimates of effective field size to square field sizes is the equivalent square table published in the British Journal of Radiology.²⁵ Estimating the effective field size equivalent to an irregular field is best handled via irregular field calculations as discussed later in this chapter.^26,27

MU Calculations for Asymmetric X-Ray Collimators

Asymmetric x-ray collimators (also referred to as independent jaws) allow independent movement of an individual jaw and may be available for one jaw pair or both pairs. Because MU calculations and treatment planning methods generally rely on symmetric jaw data, the dosimetric effects for asymmetric jaws must be fully documented before being implemented in the clinic. Several investigators have examined the effects of asymmetric jaws on PDD,collimator scatter, and isodose distributions.^28,29 Monitor unit calculations for asymmetric jaws are only slightly more complex than for symmetric jaws.²² Typically, one simply applies an off-axis ratio (OAR) or off-center ratio (OCR) correction factor that depends only on the distance from the machine’s central axis to the center of the independently collimated open field.^30,31 PDD is only minimally affected, but isodose curve shape can be altered and must be investigated for the particular treatment unit. Calculations for asymmetric wedge fields follow similar procedures by simply incorporating a wedge OAR or OCR.^29,32

MU Calculations for Multileaf Collimator

Multileaf collimators (MLCs) have nearly completely replaced conventional alloy field shaping for photon beams in most clinics around the world. Several investigators have examined the effects of the Varian MLC design (tertiary system) on PDD, collimator scatter, and isodose distributions.³³ The effects due to field area shaped by this type of MLC on PDD and beam output parameters are similar to those resulting from Cerrobend field shaping. Thus, the dose/MU calculation methods discussed previously apply by simply using the equivalent area defined by the MLC. The collimator scatter factor and the dose in free space are determined using the x-ray collimator jaw settings, with an off-axis factor applied for any asymmetric jaw settings.

It should be noted, however, that for MLC systems that replace one of the collimating jaws, the MLC field shape can be a determining factor in selecting the appropriate output factor.³⁴ For example, in the case of the Elekta linacs (Elekta AB, Crawley, United Kingdom), in which the lower jaws are replaced by the MLC system, the calculation takes into consideration the collective blocked area that is created by both the MLC leaves and the lower backup diaphragms.³⁵ For an MLC system that replaces the upper jaw (e.g., Siemens linac, Siemens Medical Solutions USA, Inc., Malvern, PA), Das et al.³⁶ describe a method that relies on the blocked area for determining all the calculation parameters (mainly output, percent depth dose, and scatter factor).

Hence, because of MLC design differences and the fact that vendors are continuing to modify/improve MLC designs, the authors caution that when a new linac is installed, the impact of the MLC on the institution’s MU calculation procedure should be fully documented before clinical use. The reader is also encouraged to review the AAPM Task Group 50 Report, which provides more detail on various MLC types and discusses quality assurance (QA) and MU calculations.³⁷

FIGURE 7.2. Outline of mantle field illustrating method of determining scatter-to-air ratio, used for irregular-field dose calculations. (From Cundiff JH, Cunningham JR, Golden R, et al. A method for the calculation of dose in the radiation treatment of Hodgkin’s disease. Am J Roentgenol 1973;117:30–34.)

MU Calculations for Irregular Fields

For large, irregularly shaped fields and at points off the central axis, it is necessary to take account of the off-axis change in intensity (relative to the central axis) of the beam, the variation of the SSD within the field of treatment, the influence of the primary collimator on the output factor, and the scatter contribution to the dose. Changes in the beam quality as a function of position in the radiation field also should be considered.^38,39

The general method used for irregular-field calculations consists of summation at each point of interest of the primary and scatter irradiation, with allowance for the off-axis change in intensity (off-axis factor) and SSD.^26,27 The MUs required to deliver a specified tumor dose at an arbitrary point in an irregular field (Fig. 7.2) can be calculated as follows:

where the parameters used are:

Computer implementations vary, but typically include using the expanded field size at a depth for the SAR calculation, determining the off-axis factor using the distance from the central axis to the slant projection of the point of calculation to the SSD plane along a ray from the source, and determining the zero-area TAR using the slant depth along a ray going from the source to the point of calculation. It is generally accepted that the off-axis factor should be multiplied by the sum of the zero-area TAR and the SAR as originally proposed.

Beam quality is a function of position in the field for beams generated by linear accelerators.^38,39 The TAR₀ may be expressed as a function of position in the beam so that changes in beam quality can be incorporated into calculations, and it can be related to the half-value layer (HVL) of water by the following equation:

where d is the depth of the point of reference, d_max is the depth of maximum dose, r is the radial distance from the central axis of the beam to the point of calculation, e is the base of the natural logarithm, and HVL(r) is the beam quality expressed as the HVL measured in water.

MU Calculations for Rotation Therapy

The classic method for MU calculations for rotation therapy is given by the following equation:

and the MU per degree setting is given by:

where the symbols have the previous meaning and TAR_avg is an average TAR (averaged over radii [depth in the patient] at selected angular intervals, such as 10 or 20 degrees).

Most recently, manufacturers of linacs and their associated planning systems have introduced features that provide rotational intensity modulated radiation therapy (IMRT) capability^40,41 (e.g., Elekta VMAT⁴² and Varian RapidArc⁴³). The linac-based rotational IMRT concept was first proposed by Yu^44,45 and called intensity modulated arc therapy (IMAT), but planning software was not commercially available at that time. Rotational IMRT approaches on conventional linacs may provide even more conformal dose distributions delivered in a shorter treatment time, compared with SMLC-IMRT (step and shoot) or DMLC-IMRT (dynamic) approaches that use only a limited number of gantry directions. In addition, plan optimization is simpler because it eliminates the planner’s iterative choices of beam number and direction. The conventional MLC approach for rotational IMRT is likely to improve IMRT plan quality and delivery efficiency,⁴³ although this remains somewhat controversial at present^46,47 and more users will need to report their rotational IMRT experiences over the next few years. Monitor unit calculations for this more complex form of rotational therapy are generated via advanced treatment planning systems having this capability. All are techniques in which the MLC shape changes during a rotation therapy, and depending on the specific linac, other parameters, such as dose rate, may also change. This advanced type of treatment delivery is currently checked via phantom measurements rather than manual calculations prior to the patient’s treatment.

It is apparent that the simple MU manual calculations methods described here are no longer adequate for the complex technologies in use today. Modern computer plans utilize dose weight points of interest, which classical MU calculation methods may not address; in addition, geometries that include the presence of heterogeneities, added tertiary devices such as MLCs and asymmetric jaws, and beam intensity modulation can all be problematic for manual check calculations. For example, when the dose weight point is within a small volume of mass surrounded by low-density tissue (e.g., a coin lesion in lung) or one that is at the border of a chest wall and lung (as in a postmastectomy patient), MU calculations cannot easily be confirmed by simple hand-calculated MU methods, and one must rely on the planning system for calculations such as these, emphasizing the importance of fully testing such system prior to clinical use. Dedicated commercial software for MU verification is now available, for example, RadCalc (LifeLine Software Inc, Austin, TX) based on the work of Kung et al.⁴⁸ and IMSure (Standard Imaging, Middleton, WI) based on the work of Yang et al.⁴⁹ Obviously such systems must also be validated by the physics user prior to clinical use.

FIGURE 7.3. Typical photon and electron beam central-axis percentage depth dose (DD) curves for a 10 × 10 cm beam for megavoltage beams ranging from ⁶⁰Co to 18-MV x-rays and 6- to 20-MeV electron beams.

TABLE 7.1 BEAM CHARACTERISTICS FOR PHOTON BEAM ENERGIES OF INTEREST IN RADIATION THERAPY

CLINICAL PHOTON BEAM DOSIMETRY

Percent Depth Dose and Single-Field Isodose Charts

The central-axis PDD expresses the penetrability of a radiation beam. Table 7.1 summarizes beam characteristics for x-ray and γ-ray beams typically used in radiation therapy and lists the depth at which the dose is maximum (100%) and the 10-cm depth PDD value. Representative PDD curves are shown in Figure 7.3 for conventional SSDs. As a rule of thumb, an 18-MV, 6-MV, and ⁶⁰Co photon beam loses approximately 2%, 3.5%, and 4.5% per centimeter, respectively, beyond the depth of maximum dose, d_max (values are for a 10 × 10 cm field, 100-cm SSD). There is no agreement as to what is the single optimal x-ray beam energy; instead, institutional bias or radiation oncologist training typically influences its selection, and it is usually treatment site specific. As pointed out in Chapter 5, most modern linacs are multimodality, and provide a range of photon and electron beam energies ranging from 4 to 25 MV, with 6- and 15 or 18-MV x-ray beams the most common.

Isodose charts provide much more information about the radiation beam characteristics than do central-axis PDD data alone. However, even isodose charts are limited in that they represent the dose distribution in only one plane (typically the one containing the beam’s central axis) and are usually available only for square or rectangular fields. Isodose charts are usually measured in a water phantom with the radiation beam directed perpendicular to the phantom’s flat surface. Isodose curves show the relative uniformity of the beams across the field at various depths, and also provide a graphical depiction of the width of the beam’s penumbra region. ⁶⁰Co teletherapy units exhibit a relatively large penumbra, and their isodose distributions are more rounded than those from linac x-ray beams. This is due to the relatively large source size (typically 1 to 2 cm in diameter vs. only a few millimeters for linacs). Linac beam penumbra width does increase slightly as a function of energy and if unfocused MLC leaves are used, but is still much less than that for ⁶⁰Co units. In addition to the smaller penumbra, linac x-ray isodose distributions have relatively flat isodose curves at depth. However, at shallow depths, particularly at d_max, linac x-ray beams typically exhibit an increase in beam intensity away from the central axis; this beam characteristic is referred to as the dose profile horns and depends on flattening filter design. In general, each treatment unit has unique radiation beam characteristics, and thus, isodose distributions must be measured, or at least verified, for each specific treatment unit.

Another important point to understand is how the radiation field size is defined. The radiation field size dimensions refer to the distance perpendicular to the beam’s direction of incidence that corresponds to the 50% isodose at the beam’s edge. It is defined at the skin surface for SSD treatments, and at the SAD for isocentric treatments.

FIGURE 7.4. Relative surface dose versus field size with blocking tray in place for 6- and 18-MV photons. (From Klein EE, Purdy JA. Entrance and exit dose regions for Clinac-2100 C. Int J Radiat Oncol Biol Phys 1993;27:429–435.)

FIGURE 7.5. The variation of surface dose and depth of maximum dose as a function of the angle of incidence of the x-ray beam with the surface (4 MV, 10 × 10 cm).

FIGURE 7.6. Enhancement of exit dose for (A) 6-MV and (B) 18-MV photons for a 15 × 15 cm field at 100-cm source-to-axis distance versus backscatter depth for various backscattering materials. (From Klein EE, Purdy JA. Entrance and exit dose regions for Clinac-2100 C. Int J Radiat Oncol Biol Phys1993;27:429–435.)

Depth-Dose Buildup Region

When a photon beam strikes the tissue surface, electrons are set in motion, causing the dose to increase with depth until the maximum dose is achieved at depth d_max. As the energy of the photon beam increases, the thickness of the buildup region is increased. The subcutaneous tissue-sparing effects of higher-energy x-rays, combined with their great penetrability, make them well suited for treating deep lesions. In general, the dose to the surface and in the buildup region for megavoltage photon beams generally increases with increasing field size and with the insertion of blocking trays made of plastic or other type material in the beam (Fig. 7.4). The blocking trays should be at least 20 cm above the skin surface because skin doses are significantly increased for lesser distances. Copper, lead, or lead glass filters beneath the blocking tray can be used to remove the undesired lower-energy electrons that contribute to skin dose, but this is nowadays rarely done routinely in the clinic.^50,51

As the angle of the incident radiation beam becomes more oblique, the surface dose increases, and d_max moves toward the surface (Fig. 7.5). This is primarily due to more secondary electrons’ contribution from the media below the surface along the oblique path of the beam.⁵²

Depth Dose/Exit Dose Region

The skin and superficial tissue on the side of the patient from which the beam exits receive a reduced dose if there is insufficient backscatter material present. The amount of dose reduction is a function of x-ray beam energy, field size, and the thickness of tissue that the beam has penetrated reaching the exit surface. For a 6-MV beam, a 15% reduction in dose with little dependency on field size has been reported,⁵⁰ and for 18-MV beams, an 11% reduction in exit dose was measured.⁵³ In general, the addition of a thickness of tissue-equivalent material on the exit side equivalent in thickness to approximately two-thirds of the d_max depth is sufficient to provide full dose to the build-down region on the exit side. Figure 7.6 shows the effects of various backscattering media when placed directly behind the exit surface.

Tissue Heterogeneities and Tissue Interface Dosimetry

The presence of tissue heterogeneities, such as air cavities, lungs, bony structures, and prostheses, can greatly impact the calculated dose distribution. The change in dose is due to the perturbation of the transport of primary and scattered photons and that of the secondary electrons set in motion from photon interactions. Depending on the energy of the photon beam and the shape, size, and constituents of the inhomogeneities, the resultant change in dose can be large.

Perturbation of photon transport is more noticeable for lower-energy beams. There is usually an increase in transmission, and therefore dose, when the beam traverses a low-density inhomogeneity. The reverse applies when the inhomogeneity has a density higher than that of water. However, the change in dose is complicated by the concomitant decrease or increase in the scatter dose. For a modest lung thickness of 10 cm, there will be about a 15% increase in the dose to the lung for a ⁶⁰Co or 6-MV x-ray beam, but only about 5% for an 18-MV x-ray beam.⁵⁴

When there is a net imbalance of electrons leaving and entering the region near an inhomogeneity (interfaces of different media), the condition of electron equilibrium is disrupted. The dose distribution in the patient in such transition zones depends on radiation field size (scatter influence), distance between interfaces (e.g., air cavities), differences between physical densities and atomic number of the interfacing media, and the size and shape of the different media. Because electrons have finite travel, the resultant change in dose is usually local to the vicinity of the inhomogeneity but may be quite large. The effects are more noticeable for the higher photon energy beams due to the increased energy and range of the scattered electrons. Near the edge of the lungs and air cavities, the reduction in dose can be larger than 15%.⁵⁵

For inhomogeneities with density larger than water, there will be an increase in dose locally due to the generation of more electrons. However, most dense inhomogeneities have atomic numbers higher than that of water so that the resultant dose perturbation is further compounded by the perturbation of the multiple coulomb scattering of the electrons. Near the interface between a bony structure and waterlike tissue, large hot and cold dose spots can be present. Several benchmark measurements have been reported for various geometries simulating clinical situations and are discussed briefly in the following sections.

Air Cavities

Air cavities that appear in various locations of the body, most particularly in the head and neck region, pose a problem due to loss of equilibrium at the air–tissue boundaries internal to the patient. Epp et al.⁵⁶ reported that for cobalt beams, a reduction in dose of approximately 12% was found for a typical larynx air cavity, which recovered within 5 mm in the new buildup region. The loss was due to a lack of forward scattered electrons. Epp et al.⁵⁷ reported that for a 10-MV x-ray beam, a 14.5% loss was measured at the distal interface of the air cavity with a buildup curve that plateaued within 20 mm of the interface. Klein et al.⁵⁸ measured distributions about air cavities for 4-MV and 15-MV x-ray beams in both the distal and proximal regions. The combined dose distribution in a parallel-opposed fashion showed a 10% loss at the interfaces for both beam energies.

Lung Tissue

Although the problem of reestablishing equilibrium for lung interfaces is not as severe as with air cavities, a transition zone region at the lung–tissue interface still exists over the range of typical clinical photon beam energies. Rice et al. measured responses within various simulated lung media for 4-MV and 15-MV x-rays using a parallel-plate ion chamber and a phantom constructed of solid water and simulated lung material (average lung material density, ρ = 0.31 g/cm³; some additional measurements made with materials having densities of 0.015 g/cm³ and 0.18 g/cm³).⁵⁹ Figure 7.7 shows the results in terms of measured CFs for the 15-MV beam. A considerable buildup curve was observed (10% change in CF) for small fields (5 × 5 cm²) for the 15-MV beam, which began in the distal region of the lung and plateaued about 5 cm beyond the simulated lung interface.

Bone–Soft Tissue Interfaces

Das et al.⁶⁰ measured dose perturbation factors (DPFs) proximal and distal for simulated bone–tissue interface regions using a parallel-plate chamber for both 6- and 24-MV x-ray beams. They reported DPFs of 1.1 for the 6-MV beam and 1.07 for the 24-MV beam at the proximal interface. At the distal interface, a DPF of 1.07 was measured for the 24-MV beam, whereas the 6-MV beam exhibited a DPF of 0.95, resulting in a new buildup region in soft tissue. Note, both buildup and build-down regions dissipate within a few millimeters from the interfaces and the perturbations are independent of thickness and lateral extent of the bone or radiation field size.

Metal Prostheses

Das and associates measured DPFs following a 10.5-mm-thick stainless steel layer simulating a hip prosthesis geometry.⁶⁰ They reported a DPF of 1.19 for 24-MV photons, but only 1.03 for 6-MV photons; on the proximal side, they reported a DPF of 1.30 due to the backscattered electrons that was independent of energy, field size, or lateral extent of the steel. These interface effects dissipated within a few millimeters in polystyrene. Other reports dealing with dosimetry perturbations due to metal objects are included in the references.^61,62

Niroomand-Rad et al. reported on dose perturbation effects at the tissue–titanium alloy implant interfaces in patients with head and neck cancer treated with 6-MV and 10-MV photon beams.⁶³ They found at the upper surface (toward the source) of the tissue–dental implant interface DPFs of 1.22 and 1.20 for the 6-MV and 10-MV photon beams, respectively. At the lower interface, dose reduction was approximately −13.5% and −9.5% for the 6-MV and 10-MV beams, respectively.

The most complete information currently available on hip prosthesis dosimetry is found in AAPM Report 81 (TG 63).⁶⁴ The report provides the current state of scientific understanding and clinical dosimetry in use for patients with high-Z hip prostheses undergoing radiation therapy. Beam arrangements that avoid the prosthesis should always be a first consideration. If this cannot be done, valuable information is available in Report 81, including values for different prostheses’ electron density, approximate attenuation of the beam passing through the prosthesis, and possible dose increase to the hip bone. It should also be noted that some of the data provided and recommendations are also applicable to patients having other implanted high-Z prosthetic devices such as pins and humeral head replacements.

Silicone–Soft Tissue Interfaces

Klein and Kuske reported on interface perturbations with silicon breast prostheses.⁶⁵ Such prostheses have a density similar to breast tissue but have a different atomic number. They observed a 6% enhancement at the proximal interface and a 9% loss at the distal interface.

Wedge Filter Dosimetry

When a wedge filter is inserted into the beam, the dose distribution is angled at some specified depth to some desired angle relative to the incident beam direction over the entire transverse dimension of the radiation beam (Fig. 7.8). For cobalt units, the depth of the 50% isodose usually is selected for specification of the wedge angle, whereas for high-energy linacs, higher-percentile isodose curves, such as the 80% curve, or the isodose curves at a specific depth (10 cm) are used to define the wedge angle.

Linacs are typically equipped with multiple wedges that may be used with an allowed range of field sizes. Although linac wedges can be designed for any desired wedge angle, 15-, 30-, 45-, and 60-degree wedges are the most common.

FIGURE 7.7. A: Dose correction factors as a function of depth for a transition zone geometry that simulates a lung–tissue interface for three different field sizes and a lung thickness of 10 cm for 15-MV x-rays. The modification to the primary dose only on the central axis (shown by the dashed curve) is independent of field size. B: Dose correction factors as a function of depth for a transition zone geometry that simulates a lung–tissue interface for three different densities, a 5 × 5 cm field, and a lung thickness of 10 cm for 15-MV x-rays. The modification to the primary dose only on the central axis is shown by the dashed curve. (From Rice RK, Mijnheer BJ, Chin LM. Benchmark measurements for lung dose corrections for x-ray beams. Int J Radiat Oncol Biol Phys 1988;15:399–409.)

FIGURE 7.8. Isodose distributions for a 6-MV x-ray beam with an 8 × 8 cm field size. A: Open field. B: Field with a 45-degree wedge. (From Khan FM. The physics of radiation therapy, 2nd ed. Baltimore: Williams & Wilkins, 1994.)

FIGURE 7.9. Parameters of the wedge beams: φ is the wedge angle, θ is the hinge angle, and S is separation. Isodose curves for each wedge field are parallel to the bisector. (From Khan FM. The physics of radiation therapy, 2nd ed. Baltimore: Williams & Wilkins, 1994.)

Some linacs (Elekta AB, Sweden) feature a single wedge, referred to as a universal wedge, located in the treatment head, and the desired wedged dose distribution is obtained by the proper combination of wedged and unwedged treatment. A simple approximate model for combining open and wedged fields was first proposed by Tatcher,⁶⁶ in which the effective wedge angle θ_E, resulting by the addition of a wedged and unwedged beam, is equal to the nominal wedge angle θ_W for the wedged beam, weighted by the fraction of wedged field B:

The Philips Medical Systems Division⁶⁷ proposed a slightly different method as follows:

Petti and Siddon⁶⁸ investigated both methods and showed that these are approximations to an exact theoretical solution, which is given by:

Most importantly, their investigations showed that Tatcher’s approximation is good only for values of θ_W less than 45 degrees, and thus is inadequate for accelerators such as Elekta, which use a 60-degree motorized universal wedge. They did show that for the field sizes studied (up to 20 × 20 cm), the Philips relationship was valid to within 3 degrees.

The wedged isodose curves can be normalized in two different ways. In some older systems, the wedge dose distributions have the wedge factor (i.e., the ratio of the measured central-axis dose rate with and without the wedge in place) incorporated into the wedged isodose distribution. More commonly, the wedge isodose curves are normalized to 100% at d_max, and a separate wedge factor is used to calculate the actual treatment MUs or time. McCullough et al.⁶⁹ noted that wedge factors measured at d_max usually are accurate to within 2% for depths up to 10 cm, but at greater depths can be inaccurate to 5% or more. The inclusion (or noninclusion) of the wedge factor is an extremely important point to understand because serious error in dose delivered to the patient can occur if used improperly.

Sewchand et al.⁷⁰ and Abrath and Purdy⁷¹ pointed out that beam hardening results when a wedge is inserted into the radiation beam. The PDD, therefore, can be considerably increased at depth. Differences reported were nearly 7% for a 4-MV 60-degree wedge field PDD from the open field PDDs at 12-cm depth, and a 3% difference in depth-dose values between the wedge field and the open field for a 60-degree wedge using 25-MV x-rays was reported.

Modern computer-controlled medical linacs now have software features that allow the user to create a wedge-shaped dose distribution by moving one collimator jaw across the field in conjunction with adjustment of the dose rate over the course of the daily single-field treatment.⁷² This technology provides superior dose distributions and eliminates the previously mentioned beam-hardening problem seen in physical wedges. This feature can deliver a greater number of wedge angles, and over larger field sizes, including asymmetric field sizes (30 cm in the wedge direction, with 20 cm toward the wedge “heel,” and 10 cm toward the wedge “toe”). The increased number of angles enhances planning options but also complicates commissioning and QA.⁷³

When the patient’s treatment is planned, wedged fields are commonly arranged such that the angle between the beams, the hinge angle (θ), is related to the wedge angle (φ) by the following relationship (Fig. 7.9):

For example, as shown in Figure 7.10, 45-degree wedge fields orthogonal to one another yield a uniform dose distribution.

TREATMENT PLANNING: COMBINATION OF TREATMENT FIELDS

Parallel-Opposed Fields

When only two unmodified x-ray beams are used in radiation therapy, they usually are parallel-opposed beams (i.e., directed toward each other from opposite sides of the anatomic site with the central axes coinciding). Figure 7.11 presents the normalized relative axis dose profiles from parallel-opposed photon beams for a 10 × 10 cm field at an SSD of 100 cm and for patient diameters of 15 to 30 cm in 5-cm increments. The weight of a beam denotes a numeric value assigned to the beam at some normalization point. For SSD beams, the weight specifies the relative dose assigned to the beam at d_max, and for isocentric beams, at isocenter. The beams shown are weighted 1 to 1 (i.e., assigned equal value 100% at d_max), and the dose profiles have been normalized to the cumulative midline PDD.

The maximum patient diameter easily treated with parallel-opposed beams for a midplane tumor requiring 50 Gy or less with low-energy megavoltage beams is approximately 18 cm. For “thicker” patients, higher x-ray energies produce improved dose profiles with less dose variation along the central axis without resorting to more complex multibeam arrangements.

For some treatment sites, the underdosing achieved near the skin surface with very–high-energy, parallel-opposed x-ray beams is a highly advantageous feature, but in others it may be desirable to achieve a higher dose nearer to the skin. With very–high-energy x-ray beams traversing small anatomic thicknesses, the exit dose can exceed the entry dose, and the exact dose distribution in the regions beneath the entry and exit surfaces from parallel-opposed high-energy x-ray beams must be carefully evaluated to consider properly the contribution from both entrance and exit components.

Unequal beam weightings are advantageous if the target volume is not midline. Figure 7.12 shows normalized central-axis dose profiles for other weightings, such as 2 to 1 and 3 to 1. The greater the unequal weighting, the greater will be the shift of the higher-dose region toward one surface and away from midline. Although in some anatomic sites unequal weighting may be advantageous, special attention must be directed to the anatomic structures in the high-dose volume.

FIGURE 7.10. Isodose distribution for two angled beams. A: Without wedges. B: With wedges. Both: 4 MV; field size, 10 × 10 cm; source-to-skin distance, 100 cm; wedge angle, 45 degrees. (From Khan FM. The physics of radiation therapy, 2nd ed. Baltimore: Williams & Wilkins, 1994.)

FIGURE 7.11. Relative central-axis dose profiles as a function of x-ray energy (⁶⁰Co or 4, 6, 10, and 25 MV) and patient thickness (15, 20, 25, and 30 cm). The parallel-opposed beams are equally weighted, and the profiles are normalized to unity at midline. Because of symmetry, only half of each profile is shown.

Multiple-Beam Arrangements

Figure 7.13 shows three commonly used coaxial three-field beam arrangements. A direct anterior field with two anterior oblique fields can be used to generate a high-dose region where the three fields overlap, whereas a low-dose region exists beyond this intersection point. For example, if this arrangement is used for treating the mediastinum, the spinal cord might be included in the anterior beam but spared by the anterior oblique beams. Moving the anterior oblique fields laterally to form a parallel-opposed pair yields a rectangular isodose region with a more uniform dose gradient; however, the magnitude of the dose gradient is determined by the relative weighting of the beams and the thickness of tissue traversed. An anterior field with two symmetrically placed posterior oblique beams yields elongated isodose curves. The degree of elongation is determined by the relative thickness of tissue each beam traverses to the point of intersection and by the relative weights of the beams. Three-field arrangements are often useful for treating tumors lateral to the midline of a patient.

Three-field nonaxial (noncoplanar) arrangements are readily achieved with linacs by rotating the table and gantry. A common technique for treating pituitary tumors uses two lateral fields and a vertex field with the beam entering through the top of the head. Astrocytomas often are treated with parallel-opposed lateral fields and a frontal field entering through the forehead. A 90-degree couch rotation is used with the gantry rotated laterally for the vertex or frontal fields. The lateral fields are also rotated by collimator to ensure the “heels” of the wedges are in the plane of the vertex/frontal field trajectory.

Four-field techniques typically are used in such sites as the abdomen or the pelvis. In most instances, the arrangements consist of pairs of parallel-opposed fields, with a common intersecting point, which yield a “boxlike” isodose distribution. Figure 7.14 compares the dose distributions achieved with a four-field “box technique” for 6- and 18-MV x-ray beams. The central dose distribution is similar for all beam energies, but the greater penetrability of the higher-energy beams yields a lower dose to the region outside the box. Variations in the dose gradient are achieved by differential weighting of each pair of beams. Figure 7.15 shows other possible four-beam arrangements. Angulation of the beams yields a diamond-shaped dose distribution. A butterfly-shaped distribution is achieved if each pair of beams has a point of intersection lying on a common line but separated by a few centimeters.

Treatments involving more than four gantry angles, historically required with orthovoltage x-ray units to treat deep, midline lesions, were originally rarely used with high-energy megavoltage therapy units. However, with the broad introduction of three-dimensional conformal radiation therapy (3DCRT) and IMRT, there has been an increase in multibeam treatments such as the 3DCRT six-field technique commonly used for the treatment of prostate carcinoma⁷⁴ and the nine-field technique commonly used for head and neck cancer IMRT treatments.⁷⁵ More recently, these multibeam arrangements are further refined by adding segments of field with the same beam angle to either improve dose homogeneity or to intentionally generate an inhomogeneous dose distribution (e.g., the simultaneous integrated boost technique).⁷⁶

FIGURE 7.12. Dose profiles achieved with unequal weightings of parallel-opposed photon beams; profiles are normalized to unity at midline.

FIGURE 7.13. Three-field coaxial beam arrangements: Dose distribution for two different beam arrangements using 6-MV x-ray beams, 8 × 10 cm field size, 100-cm source-to-skin distance. Isodose curves have been renormalized to show the 100% line almost encompassing the target volume. A: Anterior field with two anterior oblique fields at 40 degrees off the midline, all equally weighted. B: Anterior field with a weight of 0.8 with two equally weighted (1) posterior oblique fields separated by 120 degrees.

Rotation Therapy

Rotational (or arc) therapy techniques, in which the treatment is delivered while the gantry (and thus the radiation beam) rotates around the patient, can be thought of as an infinite extension of the multiple-field techniques already described. This technique is most useful when applied to small, symmetric, deep-seated tumors, and usually is limited to field sizes less than approximately 10 cm in width for the treatment of centrally located lesions (i.e., there is approximately an equal amount of tissue in all directions around the lesion).

Dose distributions generated by rotational techniques are not very sensitive to the energy of the photon beam. Figure 7.16 illustrates this fact, showing the dose distribution achieved using a 6-MV x-ray beam, and also the distribution using an 18-MV x-ray beam. There is a little less elongation in the direction of the shorter dimension of the patient’s anatomy for the 18-MV beam, and the dose distribution in the periphery is slightly lower.

In arc therapy techniques, one or more sectors of a 360-degree rotation are skipped to reduce the dose to critical normal structures. When a sector is skipped, the high-dose region is shifted away from the skipped region. Therefore, the isocenter must be moved toward the skipped sector; this technique is referred to as past-pointing, as illustrated in Figure 7.17.

FIGURE 7.14. Four-field “box technique” coaxial beam arrangements (equal beam weightings). A: 6-MV x-ray beams. B: 18-MV x-ray beams. Note the improved dose distribution with the higher-energy beam technique (more uniform dose in the target region and lower doses near the femoral head region of the lateral fields) as a result of the increased percentage depth for 18-MV x-rays.

FIGURE 7.15. Four-field “oblique technique” coaxial beam arrangements (6-MV x-rays, equal beam weightings). A: With common isocenter resulting in a diamond-shaped dose distribution. B: Each beam pair intersecting at two different points on a common line resulting in a butterfly-shaped isodose distribution.

The prostate, bladder, cervix, and pituitary are clinical sites that have been treated, either initially or for boost doses, with rotational or arc therapy techniques. Although the dose distributions achieved by rotation or arc therapy yield high target-volume doses, these techniques normally result in a greater volume of normal tissue being irradiated (albeit at low doses) than fixed, multiple-field techniques.

As previously indicated, modern-day rotational therapy is now delivered by a variety of techniques (e.g., tomotherapy, RapidArc, VMAT). A new technology now being implemented on linacs and one that is likely to significantly impact modern-day rotational therapy is referred to as flattening filter free (FFF); that is, the flattening filter is not present during irradiation. The FFF feature allows dose rates up to 2,400 MUs per minute, thus greatly improving efficiency in dose delivery, which is particularly important for patients receiving treatments where very high daily doses are delivered, such as stereotactic body radiation therapy (SBRT).⁷⁷ The reader is referred to the reviews by Georg et al. regarding the current status and future perspective of FFF beams.⁷⁸

FIGURE 7.16. A 360-degree rotational therapy technique. A: 6-MV x-ray beams. B: 18-MV x-ray beams. Note that there is little difference in the dose distribution when using a higher-energy beam as a result of the offsetting effects of increased percentage depth versus higher exit dose.

FIGURE 7.17. Arc therapy technique for 6-MV x-rays. A: 240-degree arc. Note that when a sector of the full 360-degree rotation is skipped, the high-dose isodose curves are shifted away from the skipped sector. B: 240-degree arc, but patient positioned so that isocenter is 2 cm lower toward the skipped sector (this technique is called past-pointing). Note high-dose isodose curves now encompass the target volume.

FIGURE 7.18. Independent or asymmetric collimators. A: Conventional symmetric pairs of collimators. B: Asymmetric collimators in which collimator jaws are allowed to move independently of each other.

FIELD SHAPING

A major constraint in the treatment of cancer using radiation is the limitation in the dose that can be delivered to the tumor because of the dose tolerance of the tissue (critical organs) surrounding or near the target volume. Shielding normal tissue and critical organs has allowed the radiation oncologist to increase the dose to the tumor volume while maintaining the dose to critical organs below some tolerance level. The frequently used tolerance doses for these organs are not absolute and depend on a number of clinical and treatment factors. Depending on the predominantly serial or parallel organization of the organ at risk, a large dose can sometimes be given to fractional volumes of organs with a parallel structure (liver, kidneys, lungs).^79,80 Shielding is usually accomplished using collimator jaws (with the asymmetric feature) and multileaf collimators, in which the beam aperture (field shape) is customized for individual patients. Use of low–melting-point alloy blocks is rapidly being replaced with the MLC technology.

MLC and Associated Dosimetry

Asymmetric Collimator Jaws

Field shaping and abutted field radiation therapy techniques have been made even more versatile with the asymmetric jaw feature found on modern-day linacs. This feature allows each set of jaws to open and close independently of each other (Fig. 7.18). The collimator jaw provides greater attenuation than the MLC leaf or alloy block, thus providing an advantage (which is readily apparent on portal films) in reducing the dose to blocked regions.

Depth-dose characteristics for asymmetric fields are similar to those of symmetric fields as long as the degree of asymmetry is not too extreme. Clinical sites where asymmetric jaws are typically used include breast (Fig. 7.19), head and neck, craniospinal, and prostate. In addition, the use of asymmetric jaws as beam splitters, for field reductions, and with MLCs is helpful for most sites. Several authors have reported on the use of asymmetric jaws to match supraclavicular and tangential fields for breast irradiation.^81–83 Such technology allows a single setup point for all of the treatment fields, including the posterior axillary field. The Y-jaws can beam split the caudal and cephalic regions for the supraclavicular and tangential beams, respectively, and the X-jaws are used to shield the ipsilateral lung and contralateral breast. Hence, a common match plane with one common isocenter can be used for all portals, eliminating the need to move the patient between portals, thus reducing overall patient setup time by almost a factor of two. In addition, the increased attenuation by the jaws reduces the dose to the contralateral breast and lung.⁸⁴ A technique for matching lateral head and neck fields and the supraclavicular field using asymmetric jaws was described by Sohn et al.⁸⁵

Multileaf Collimation

Multileaf collimation, first introduced in Japan in the 1960s,⁸⁶ has now gained widespread acceptance and has replaced alloy blocking as the standard of practice for field shaping in modern radiation therapy clinics. The different manufacturers’ MLC systems vary with respect to MLC location, leaf design, and field size coverage. The leaves are typically carried on two opposed carriages that transport the leaves in unison. The leaves have individual controls that are computer assigned and positioned. Initially, most commercial MLC systems were designed to serve as a block replacement, but now provide for dynamic IMRT delivery as well.

Elekta first introduced its MLC system in the late 1980s.⁸⁷ Their current MLC system replaces the upper photon collimator jaws, and therefore, the maximum field size can open to a full 40 × 40 cm. The MLC system is augmented by parallel diaphragms, which increase the leaf’s attenuation by an additional two HVLs.

The Varian MLC system is considered a tertiary system placed below the photon collimator jaws. The latest Varian MLC (non-SRS) is a 120-leaf (60 on each side) system, in which the middle 20 cm consists of 0.5-cm-wide leaves, while the outer 20-cm leaves still project to 1.0-cm widths. This set of leaves projects to 16.0 cm in length at isocenter, and the leaf span range (maximum–minimum positions on the same carriage) is limited to 14.5 cm. The leaves move perpendicular to the beam’s central axis. The distance from the x-ray target to the bottom of the leaves (on the central axis) is 54.0 cm. The leaves fan away from the central axis so that their sides are divergent with the beam’s fan lines. The leaves are interdigitated by a tongue-and-groove design. The reader is referred to the references for details of the Siemens MLC system in which the lower collimating jaws are replaced with a double-focused leaf system.³⁶

For the Varian MLC system, leaf transmission values of 1.5% to 2.0% for a 6-MV beam and 1.5% to 5% for an 18-MV beam have been reported.^33,88 These values are lower than those found for alloy blocks (3.5%), but higher than those for collimator jaw transmission (that being <1.0%). Transmission through abutted (closed) leaf pairs was as high as 28% for 18-MV photons on the central axis. The abutment transmission decreased as a function of off-axis distance to as low as 12%.

Figure 7.20 shows a comparison of MLCs and alloy blocks regarding penumbra. The discrete steps of the MLC systems introduce undulations in the isodose lines. This effect causes an apparent increase in penumbra with wave patterns after the undulations. Single, focused MLC systems have a slightly larger penumbra than do alloy shields and have an even larger difference in comparison with collimator jaws. Boyer et al. found the penumbra (80% to 20% isodose lines) generated by leaf ends to be wider than those generated by upper collimator jaws by 1.0 to 1.5 mm, and 1.0 to 2.5 mm compared with the lower jaws, depending on energy and field size.⁸⁹ Powlis and associates compared multileaf collimation and alloy field shaping and found few differences.⁹⁰ LoSasso and Kutcher found similar results and concluded that geometric accuracy is even improved with MLCs.⁹¹

The penumbras measured for the leaf sides are comparable with those found for upper jaws due to their divergent nature. The penumbra increase and stair-stepping effect are most prominent at d_max. The effects diminish at depth due to the influence of scattered electrons and photons as the scatter-to-primary ratio increases with depth. Adding an opposed beam leads to further smoothing of the undulations and penumbra differences become less significant. For multibeam arrangements, the differences in dose distribution between MLC and alloy shields are negligible.

Two methods for designing the optimal MLC configurations to fit the treatment plan’s field apertures have evolved: (a) configuring the MLC based on a digitized film image using a dedicated MLC workstation (with or without automated optimization), and (b) configuring the MLC using treatment planning system software. The main limitation in optimizing the MLC leaf settings to conform to the shaped field is the discrete leaf steps. Most field shapes require only minor adjustment of collimator angle to achieve minimal discrepancy between the desired and resultant field shape. The criteria for optimizing the MLC leaf settings are governed by placing the most leaf ends tangent to the field and also maintaining the same internal area as originally prescribed. MLC shaping systems typically provide an option to place the leaf ends entirely outside the field (exterior), entirely within the field (interior), or crossing the field at midleaf (leaf-center insertion). The last is the most widely used criterion because the desired field area is more closely maintained. However, this choice leads to regions in which some treatment areas are shielded and some normal tissues are irradiated. Zhu et al. reported on a variable insertion technique in which leaves are placed only far enough into the field to cause the 50% isodose contour to undulate outside and up to the desired contour.⁹²LoSasso et al. reported on a method in which each leaf is inserted such that the treatment area covered by the leaf equals the normal tissue area that is not spared.⁹³ Brahme also demonstrated optimal choices for choosing a collimator angle to optimize leaf direction, depending on whether the field shape is convex or concave.⁹⁴ Du et al. reported on a method that defines optimal leaf positioning in combination with optimal collimator angulation.⁹⁵ Typically, the optimal direction for the leaf motion is along the narrower axis. For a simple ellipse the optimal leaf direction is parallel to the short axis. Reports on the effects of tissue heterogeneities on penumbra and resultant field definition indicate that the penumbra in lung increases (especially for 18-MV photons), whereas in bone, it decreases for both alloy blocks and MLCs.³³

As indicated previously, because MLC systems are still evolving, a careful evaluation of the effect of MLCs on monitor unit calculations must be performed before clinical use. Extensive testing over the clinical range of field sizes and shapes should be undertaken before the MLC system is used clinically.

FIGURE 7.19. Treatment technique for breast cancer using independent collimators. (From Klein EE, Taylor M, Michaletz-Lorenz M, et al. A mono isocentric technique for breast and regional nodal therapy using dual asymmetric jaws. Int J Radiat Oncol Biol Phys 1994;28:753–760.)

FIGURE 7.20. Comparison of beam’s-eye-view isodose curves at 10-cm depth for multileaf collimator (solid line) and Cerrobend-shaped (dashed line) beam apertures for 18-MV photons. (From Klein EE, Harms WB, Low DA, et al. Clinical implementation of a commercial multileaf collimator: dosimetry, networking, simulation, and quality assurance. Int J Radiat Oncol Biol Phys 1995;33:1195–1208.)

FIGURE 7.21. Composite photographs illustrating the low–melting-point alloy shielding block design and fabrication process. A: Physician defining the treatment volume on the x-ray simulator radiograph. B: Physics technician adjusting the source-to-skin distance and skin-to-film distance of a hot-wire cutter to emulate simulator geometry. C: Proper-thickness foam block aligned to the central axis of the cutter. D: Foam mold cut with hot-wire cutter. E: Foam pieces aligned and held in place using a special clamping device. Molten alloy is poured into the mold and allowed to harden. F: Examples of typical shielding blocks cast using this system. (From Purdy JA. Secondary field shaping. In: Wright AE, Boyer AL, eds. Advances in radiation therapy treatment planning. New York: American Institute of Physics, 1983.)

FIGURE 7.22. Attenuation in Lipowitz metal of x-rays produced at 2, 4, 10, and 18 MV and γ-rays from ⁶⁰Co. (From Huen A, Findley DO, Skov DD. Attenuation in Lipowitz’s metal of x-rays produced at 2, 4, 10, and 18 MV and gamma rays from cobalt-60. Med Phys1979;6:147.)

FIGURE 7.23. Schematic illustrating typical geometry used in the design of a compensator filter to account for patient’s irregularly shaped surface. SSD, source-to-skin distance; CA, central axis.

Low-Melting Alloy Blocks

Although alloy blocks are rapidly disappearing from clinical use, a short section is included in this chapter for completeness. The Lipowitz metal (Cerrobend) shielding block system was introduced by Powers et al.⁹⁶ Lipowitz metal consists of 13.3% tin, 50% bismuth, 26.7% lead, and 10% cadmium. The physical density at 20°C is 9.4 g/cm³, compared with 11.3 g/cm³ for lead. The block fabrication procedure is illustrated in Figure 7.21, and more details on using this form of field shaping can be found in the review article by Leavitt and Gibbs.⁹⁷

Specific doses to critical organs may be limited by using either a full-thickness block, usually five HVLs (3.125% transmission) or six HVLs (1.562% transmission), or a partial transmission shield, such as a single HVL (50% transmission) of shielding material. The actual dose delivered under the shielded area is usually greater than these stated transmission levels because of scatter radiation beneath the blocks from adjacent unshielded portions of the field. The scatter component of the dose increases with depth as more radiation scatters into the shielded volume beneath the block. Thus, the dose to the blocked area is a function of block material, thickness (and width), field size, and energy. Figure 7.22 shows the attenuation of Lipowitz metal of x-rays produced at 2, 4, 10, and 18 MeV and ⁶⁰Co γ-rays.⁹⁸ Alloy blocks made from the standard thickness (7.6 cm) of foam molds reduce the primary beam intensity to 5% of its unattenuated value. Increasing the block thickness usually is not worthwhile because it makes the block heavier, whereas the scatter radiation contributes an equal or greater share of the dose under the blocks.

Due to the advent of multileaf collimation, metal blocks for photon beams are rarely used, as multileaf collimation afforded there to be no room for entry between treatment fields. In addition, the construction of blocks was time consuming and expensive. Also, the materials themselves, as they were heated, gave off potentially toxic air, particularly due to the lead and cadmium. The one advantage of blocks is that they provide smooth boundaries by having a continuous shape around the field and incur no field size limitation. However, the weight of the blocks can be excessive, sometimes up to 15 pounds, leading to the potential for injury to therapists and patients should they be mishandled.

COMPENSATING FILTERS

The compensating filter, introduced by Ellis et al.,⁹⁹ counteracts the effects caused by variations in patient surface curvature while still preserving the desirable skin-sparing feature of megavoltage photon beams. This is accomplished by placing the custom-designed compensating filter in the beam, sufficiently “upstream” from the patient’s surface, as illustrated in Figure 7.23. Several different compensator systems have been used in the clinic.¹⁰⁰ However, the use of physical compensators to account for patient surface curvature is almost nonexistent in clinics today due to the advent of IMRT. But one particular IMRT delivery method does use a physical compensator, which is designed from the planning system and often constructed from a third-party vendor to deliver a modulated field.¹⁰¹

BOLUS

Tissue-equivalent material placed directly on the patient’s skin surface to reduce the skin sparing of megavoltage photon beams is referred to as bolus. A tissue-equivalent bolus should have electron density, physical density, and atomic number similar to those of tissue or water and be pliable so that it conforms to the skin surface contour. Inexpensive, nearly tissue-equivalent materials used as a bolus in radiation therapy include slabs of paraffin wax, rice bags filled with soda, gauze coated with petrolatum, and synthetic-based substances, such as Super-Flab or Super Stuff.¹⁰²

Thin slabs of bolus that follow the surface contour increase the dose to the skin beneath the bolus with a maximum reduction when the bolus thickness is approximately equal to the d_max depth for the photon beam. In addition, adding bolus to fill a tissue deficit may smooth an irregular surface. A bolus also can be shaped to alter the dose distribution as well, but normally wedges are used to alter the dose distribution for megavoltage photon beams to retain skin sparing.

PATIENT POSITIONING, REGISTRATION, AND IMMOBILIZATION

Ensuring accurate daily positioning of the patient in the treatment position and reduction of patient movement during treatment is essential to deliver the prescribed dose and achieve the planned dose distribution. The reproducibility achievable in the daily positioning of a patient for treatment depends on several factors other than the anatomic site under treatment, including the patient’s age, general health, and weight. In general, obese patients and small children are the most difficult to position.

The fields to be treated typically are delineated in the computed tomography (CT) simulation process using either visible skin markings or skin markings visible only under an ultraviolet light. In some instances, external tattoos are applied. These markings are used in positioning a patient on the treatment machine using the machine’s field localization light and distance indicator and the laser alignment lights mounted in the treatment room that project transverse, coronal, and sagittal light lines (or dots) on the patient’s skin surface.

It is vital that the rigidity of the mask maintain consistency over the course of treatment. In the last 10 years the normal method of aligning the patient for treatment relied on the marks on a patient before placing a mask on. Therefore, the systems that interface with the mobilization systems and the treatment couch need to also be rigid and registered. As of now, patients are set up to treatment coordinates once they are fit on to the mobilization systems.

Numerous patient restraint and repositioning devices have been designed and used in treating specific anatomic sites. For example, the disposable foam plastic head holder provides stability for the head when the patient is in the supine position. If the patient is treated in the prone position, a face-down stabilizer can be used. This device has a foam rubber lining covered by disposable paper with an opening provided for the patient’s eyes, nose, and mouth. It allows comfort and stability as well as air access for the patient during treatment in the prone position.

A vacuum-form body immobilization system is commercially available. This system consists of a vacuum pump and an outer rubber bag filled with plastic minispheres. The rubber bag containing the minispheres is positioned to support the patient’s treatment position. A vacuum is then applied, causing the minispheres to come together to form a firm, solid support molded to the patient’s shape. The bite block (Fig. 7.24) is another device used as an aid in patient repositioning in the treatment of head and neck cancer. With this device, the patient, in the treatment position, bites into a specially prepared dental impression material layered on a fork that is attached to a supporting device. When the material hardens, the impression of the teeth is recorded. The bite-block fork is connected to a support arm, which is attached to the treatment couch, and may be used either with or without scales for registration.

Thermal plastic masks are now widely used in the United States (Fig. 7.25). A plastic sheet is placed in warm water and draped over the site, and hardens on cooling.¹⁰³ The use of thermal plastic masks allows treatments with few skin marks made on the patient because most of the reference lines can be placed on the mask. Treatments can be given through the mask; however, there is some loss of skin sparing. When skin sparing is critical, the mask may be cut out to match the treatment portal, although some of the structural rigidity is lost.

Custom molds constructed from polyurethane formed to patient contours have gained widespread use as aids in immobilization and repositioning (Fig. 7.26). The constituent chemicals for the polyurethane foam are mixed in liquid form and allowed to expand and harden around the patient while the patient is in the treatment position. These molds are used for treatment of Hodgkin’s disease with the mantle irradiation technique, in patients with cancer of the thorax or prostate, and for extremity repositioning/immobilization. Johnson et al.¹⁰⁴ reported on the effect on surface dose caused by the mold for ⁶⁰Co, 6- and 18-MV photon beams. Also, when concerned about surface dose effects caused by immobilization devices, one should not neglect understanding the effects also caused by carbon fiber couch inserts.¹⁰⁵

A bite-block system or a thermal plastic face mask system is commonly used to immobilize patients with head and neck tumors. Patients immobilized with the bite-block system typically require a larger number of adjustments than when more effective systems like the thermal face mask are used. Also, patients may prefer the face mask because most of the reference marks are on the mask rather than on the skin. However, the final assessment of accuracy and reproducibility of the daily treatment is obtained by radiographic imaging of the area treated because there is the possibility of patient movement within the mask, especially if significant tumor shrinkage or weight loss has taken place.

FIGURE 7.24. Example of a bite-block registration and immobilization system used in treatment of head and neck cancer. (Courtesy of Radiation Products Design, Buffalo, MN.)

FIGURE 7.25. Example of a registration and immobilization system (thermal plastic mask) used in treatment of head and neck cancer.

FIGURE 7.26. Examples of registration and immobilization systems (foam mold) used in treatment of the thorax. A: Mold is registered to table and the patient is registered to the mold by fiducial markings. B: Mold with cutout area to allow clear access to the treatment area, and built-in handgrip.

SEPARATION OF ADJACENT X-RAY FIELDS

Field Junctions

Different techniques for matching adjacent fields are illustrated in Figure 7.27. A commonly used gap calculation method for adjacent radiation fields is illustrated in Figure 7.28A. The separation between adjacent field edges necessary to produce junction doses similar to central-axis doses follows from the similar triangles formed by the half-field length and SSD in each field. The field edge is defined by the dose at the edge that is 50% of the dose at d_max. For two contiguous fields of lengths L₁ and L₂, the separation, S, of these two fields at the skin surface can be calculated using the following expression:

A slight modification of this formula is needed when sloping surfaces are involved, as shown in Figure 7.28B.¹⁰⁶ Typically, the skin gap location is moved a number of times to reduce the hot and cold spots that arise with this technique. Figure 7.29 illustrates the dose distribution for three different field separations.¹⁰⁷

Beam divergence may be eliminated by using a “beam splitter,” created using a five- or six-HVL block over one-half of the treatment field. The central axes of the adjacent fields, where there is no divergence, are then matched. As previously discussed, this is a useful method on linacs with the asymmetric jaws feature. Match-line wedges or penumbra generators that generate a broad penumbra for linac beams have been reported but have not found widespread use.¹⁰⁸ Here the intent is to broaden the narrow penumbra of the linacs so that it is not so difficult to match the 50% isodose levels. The resulting dose distributions are similar to those obtained with a moving gap technique. Finally, there are several reports of edge-matching techniques based on the mathematical relationships between adjacent beams and the allowed angles of the gantry, collimators, and couch.¹⁰⁹

FIGURE 7.27. Different techniques for matching adjacent fields. A: Beam’s central rays are angled slightly away from one another so that the diverging beams are parallel. B: Half-beam block to eliminate divergence. C: Penumbra generators (small wedges) to increase width of penumbra, as illustrated in D1and D2. E: Junction block over spinal cord. F: Moving gap technique. (From Bentel GC, ed. Radiation therapy planning, 2nd ed. New York: McGraw-Hill, 1996.)

FIGURE 7.28. A: Standard formula for calculating the gap at the skin surface for a given depth using similar triangles. B: Modified formula for calculating the gap for matching four fields on a sloping surface. (From Keys RA, Grigsby PW. Gapping fields on sloping surfaces. Int J Radiat Oncol Biol Phys1990;18:1183–1190.)

FIGURE 7.29. Dose distribution for geometric separation of fields with all four beams intersecting at midpoint. Adjacent field sizes: 30 × 30 cm and 15 × 15 cm; source-to-skin distance (SSD), 100 cm; anteroposterior thickness, 20 cm; 4-MV x-ray beams. A: Field separation at surface is 2.3 cm. A three-field overlap exists in this case because the fields have different sizes but the same SSD. B: The adjacent field separation increased to eliminate three-field overlap on the surface. C: Field separation adjusted to 2.7 cm to eliminate three-field overlap at the cord at 15 cm depth from anterior. (From Khan FM. The physics of radiation therapy, 2nd ed. Baltimore: Williams & Wilkins, 1994.)

FIGURE 7.30. Some solutions for the problem of overlap for orthogonal fields. A: A beam splitter, a shield that blocks half of the field, is used on the lateral and posterior fields and on the spinal cord portal to match the nondivergent edges of the beams. B: The divergence in the lateral beams may also be removed by angling the lateral beams so that their caudal edges match. Because most therapy units cannot be angled like this, the couch is rotated through small angles in opposite directions to achieve the same effect. C: A gap technique allows the posterior and lateral field to be matched at depth using a gap S on the skin surface. The dashed lines indicate projected field edges at depth D, where the orthogonal fields meet. (From Williamson TJ. A technique for matching orthogonal megavoltage fields. Int J Radiat Oncol Biol Phys 1979;5:111.)

Orthogonal Field Junctions

Figure 7.30 illustrates the geometry of matching abutting orthogonal photon beams. Such techniques are necessary, particularly in the head and neck region where the spinal cord can be in an area of beam overlap, in the treatment of medulloblastoma with multiple spinal portals and lateral brain portals. A common method of avoiding overlap is to use a half-block, as previously discussed, so that abutting anterior and lateral field edges are perpendicular to the gantry axis. In head and neck cancer, a notch in the posterior corner of the lateral oral cavity portal is commonly used to ensure overlap avoidance of the spinal cord when midline cord blocks cannot be used on anteroposterior portals irradiating the lower neck and matched to the oral cavity portals. Other techniques rotate the couch about a vertical axis to compensate for the divergence of the lateral field.¹¹⁰ The angle of rotation is given by:

Another technique is to leave a gap, S, on the anterior neck surface between the posterior field of length L and lateral field edges.¹¹¹ S can be calculated using the following formula where d is the depth of the spine beneath the posterior field:

The potential for the occurrence of radiation myelopathy resulting from a potentially excessive dose from misaligned overlapping fields is always a concern when central nervous system tumors are treated. Craniospinal irradiation is well established as a standard method of treating suprasellar dysgerminoma, pineal tumors, medulloblastomas, and other tumors involving the central nervous system. Uniform treatment of the entire craniospinal target volume is possible using separate parallel-opposed lateral cranial portals rotated so that their inferior borders match with the superior border of the spinal portal, which is treated with either one or two fields, depending on the length of the spine to be treated.¹¹² Two junctional moves are typically made at one-third and two-thirds of the total dose. The spinal field central axis is shifted away from the brain by 0.5 cm and the field size length reduced by 0.5 cm with corresponding increases in the length of the cranial field, so that a match exists between the inferior border of the brain portal and the superior border of the spine portal. To achieve the match, a collimator rotation for the whole-brain portals must be done for an angle given by the following relationship:

In addition, in order to eliminate the divergence between the cranial portal and spinal portal, the table is rotated through a floor angle given by the following relationship:

TABLE 7.2 MANAGEMENT GUIDELINES FOR RADIATION THERAPY PATIENTS WITH CARDIAC PACEMAKERS

RADIATION THERAPY PATIENTS WITH CARDIAC PACEMAKERS

As the population ages, the likelihood of encountering patients with either a pacemaker or an implantable cardioverter–defibrillator (ICD) who require radiation therapy for chest and lower neck neoplasms has become commonplace. Pacemakers are electrical devices that may stimulate either the atria, ventricles, or both (single-chamber and dual-chamber models, respectively) in order to regulate the heart’s natural rhythm.¹¹³ ICDs are generally larger than pacemakers, and also actively shock the heart to help control life-threatening, irregular heartbeats, especially those that could cause sudden cardiac arrest; most new ICDs can act as both pacemakers and defibrillators.¹¹⁴ Modern pacemakers and ICDs incorporate complementary metal oxide semiconductor (CMOS) circuitry into their generator units (encompassing a sealed lithium battery pack and circuitry only). Several groups have shown that these modern devices are much more sensitive to radiation than the older models that utilized bipolar transistor circuitry.^115–117

To safely and adequately treat such patients, it is important to understand the potential effects of radiation therapy on these devices’ operation and to take steps to minimize any actions that could jeopardize the cardiac health of the patient. The AAPM TG 34 Report provides a list of widely accepted clinical management guidelines.¹¹⁸ Potential interactions between a functioning pacemaker and the radiation therapy environment fall into two categories: (a) electromagnetic noise interference (EMI) created by the treatment machine in the course of producing high-energy photon and electron beams, and (b) damage due to radiation. Most experts now believe category 1 is no longer a source of concern. However, dose and dose rate are very much a concern.^115,117 Modern pacemakers are radiosensitive and have a significant probability of failing catastrophically at radiation doses well below normal tissue tolerance and, therefore, should never be irradiated by the direct beam. Also, several authors have shown that recommended maximum doses obtained from manufacturers have not proven to be reliable and vary greatly among manufacturers.¹¹⁹ A recent review by Hudson et al. provides the latest information and points out that radiation-induced device malfunctions are rare, and death associated with that malfunction is even more uncommon.¹¹⁹ However, they conclude that the adequacy of published guidelines is not supported by hard data. They recommend that it is important to consider all aspects of radiation therapy treatment, not just accumulated dose. These include the effect of backscatter, dose rate, fractionation, and potential EMI with new technologies such as IMRT and respiratory gating. They recommend that each radiation oncology department employ their own policy for the management of patients with pacemakers and ICDs, potentially based on an updated standard national or international guideline similar to that released by the AAPM in 1994 (Table 7.2).

FETAL DOSE

Radiation therapy is standard treatment for several malignancies (e.g., Hodgkin lymphoma, breast cancer) in which the population of women is often of childbearing age. The issues are complex, and the patient along with the radiation oncologist must evaluate treatment risk to the fetus in such cases. Radiation effects to the fetus are not fully understood and cannot be comfortably predicted for each individual case. If the decision is to irradiate, the dose levels outside the treatment fields should be quantified, and every effort should be made to lower the dose to the fetus. This may require changes in irradiation technique (i.e., modified mantle fields), elimination of double-exposure portal images, and the addition of special patient shields. The AAPM TG 36 Report provides data and techniques to estimate and reduce radiation dose to the fetus for beam energies ranging from ⁶⁰Co to 18 MV.¹²⁰ Before the pregnant patient is treated, the pregnancy stage should be known to estimate the size and location of the fetus throughout the treatment. Dose-estimation points should be selected that allow estimation of dose throughout the fetus (e.g., fundus, symphysis pubis, and umbilicus).

Two methods can be used to reduce the dose to the fetus, namely, modification of treatment techniques and the use of special shields. Modifications include changing field angle (avoiding placement of the gantry close to the fetus, i.e., treatment of a posterior field with the patient lying prone on a false tabletop), reducing field size, choosing a different radiation energy (avoiding ⁶⁰Co because of high leakage or energies of >10 MV because of neutrons), and using tertiary collimation to define the field edge nearest to the fetus. When shields are designed, the shielding device must allow for treatment fields above the diaphragm and on the lower extremities. Safety to the patient and personnel is a primary consideration in shield design. As part of the management of a pregnant patient, the treatment planning tasks listed in Table 7.3 should be performed to ensure that the dose to the fetus is kept to a minimum.¹²⁰

It should be noted that the aforementioned AAPM TG Report was based on treatment machines that predated modern MLCs. One study showed that the use of tertiary MLC systems such as those found on Varian linacs actually reduced fetal dose during irradiation of a pregnant patient.¹²¹ The leaf and carriage system provided some absorption, and if the MLC leaves were oriented in direction along the plane of concern, such as typically when a patient with Hodgkin’s disease is treated (i.e., leaves oriented along the length of the table), the reduction difference when compared to having no MLCs was on the order of 2.5 to 3.

GONADAL DOSE

Many of the reports used to estimate peripheral dose and fetal dose apply to dose estimations to both testes and ovaries.^122,123 Simultaneously, there have been studies to determine the genetically significant dose (GSD) as it applies to peripheral radiotherapy dosage to ovaries and testes. Niroomand-Rad and Cumberlin¹²⁴ combined measured data and GSD data to determine the GSD for particular treatment techniques that deliver peripheral dose to ovaries and testes. They summarized that GSDs from conventional therapies are minimal. However, there are circumstances, such as treatment of seminoma, when it is necessary to use a testicular shield to surround the testes to reduce head scatter and leakage, and some internal scatter.¹²⁵

TABLE 7.3 TREATMENT-PLANNING TASKS TO ENSURE THAT DOSE TO THE FETUS IS KEPT TO A MINIMUM

REFERENCES

1. Fischer JJ, Moulder JE. The steepness of the dose-response curve in radiation therapy. Radiology 1975;117:179–184.

2. Herring DF. The consequences of dose response curves for tumor control and normal tissue injury on the precision necessary in patient management. Laryngos 1975;85:119–125.

3. Stewart J, Jackson A. The steepness of the dose response curve for both tumor cure and normal tissue injury. Laryngoscope 1975;85:1107–1111.

4. Mackie RT, Liu HH, McCullough EC. Treatment planning algorithms: model-based photon dose calculations. In: Khan FM, ed. Treatment planning in radiation oncology, 2 nd ed. Philadelphia: Lippincott Williams & Wilkins, 2007:63–77.

5. Ahnesjö A, Aspradakis MM. Dose calculations for external photon beams in radiotherapy. Phys Med Biol 1999;44:R99–R155.

6. Khan FM, Sewchand W, Lee J, et al. Revision of tissue-maximum ratio and scatter-maximum ratio concepts for cobalt 60 and higher energy x-ray beams. Med Phys 1980;7:230–237.

7. International Commission on Radiation Units and Measurements. ICRU Report 24: determination of absorbed dose in a patient irradiated by beams of X or gamma rays in radiotherapy procedures.Washington, DC: International Commission on Radiation Units and Measurements, 1976.

8. Batho HF. Lung corrections in Cobalt 60 beam therapy. J Can Assoc Radiol 1964;15:79–83.

9. Young MEJ, Gaylord JD. Experimental tests of corrections for tissue inhomogeneities in radiotherapy. Br J Radiol 1970;43:349–355.

10. Sontag MR, Cunningham JR. Corrections to absorbed dose calculations for tissue inhomogeneities. Med Phys 1977;4:431–436.

11. Mackie TR, Scrimger JW, Battista JJ. A convolution method of calculating dose for 15-MV x-rays. Med Phys 1985;12:188–196.

12. Chetty IJ, Curran B, Cygler JE, et al. Report of the AAPM Task Group No. 105: issues associated with clinical implementation of Monte Carlo-based photon and electron external beam treatment planning. Med Phys2007;34(12):4818–4853.

13. Verhaegen F, Seuntjens J. Monte Carlo modelling of external radiotherapy photon beams (Topical Review). Phys Med Biol 2003;48(21):R107–R164.

14. Ahnesjö A, Andreo P, Brahme A. Calculation and application of point spread functions for treatment planning with high energy photon beams. Acta Oncol 1987;26:49–56.

15. Lydon JM. Photon dose calculations in homogeneous media for a treatment planning system using a collapsed cone superposition convolution algorithm. Phys Med Biol 1998;43:1813–1822.

16. Ahnesjö A. Collapsed cone convolution of radiant energy for photon dose calculation in heterogeneous media. Med Phys 1989;16:577–592.

17. Siebers JV, Keall PJ, Kawrakow I. Monte Carlo dose calculations for external beam radiation therapy. In: Dyk JV, ed. The modern technology of radiation oncology—a compendium for medical physicists and radiation oncologists, vol 2. Madison, WI: Medical Physics Publishing, 2005:91–130.

18. Cygler JE, Daskalov GM, Chan GH, et al. Evaluation of the first commercial Monte Carlo dose calculation engine for electron beam treatment planning. Med Phys 2004;31(1):142–153.

19. American Association of Physicists in Medicine. Report 85: tissue inhomogeneity corrections for megavoltage photon beams: report of Task Group 65 of the Radiation Therapy Committee. Madison, WI: Medical Physics Publishing, 2004.

20. Frank SJ, Forster KM, Stevens CW, et al. Treatment planning for lung cancer: traditional homogeneous point-dose prescription compared with heterogeneity-corrected dose-volume prescription. Int J Radiat Oncol Biol Phys2003;56(5):1308–1318.

21. Davidson SE, Popple RA, Ibbott GS, et al. Heterogeneity dose calculation accuracy in IMRT: study of the commercial treatment planning systems using an anthropomorphic thorax phantom. Med Phys2008;35:5434–5439.

22. Gibbons JP, ed. Monitor unit calculations for external photon and electron beams. Madison, WI: Advanced Medical Publishing, 2000.

23. Georg D, Huekelom S, Venselaar J. Formalisms for MU calculations, ESTRO booklet 3 versus NCS report 12. Radiother Oncol 2001;60(3):319–328.

24. Stern RL, Heaton R, Fraser MW, et al. Verification of monitor unit calculations for non-IMRT clinical radiotherapy: report of AAPM Task Group 114. Med Phys 2011;38(1):504–530.

25. Central axis depth dose data for use in radiotherapy. Br J Radiol 1996;25(Suppl).

26. Cundiff JH, Cunningham JR, Golden R, et al. A method for the calculation of dose in the radiation treatment of Hodgkin’s disease. Am J Roentgenol 1973;117:30–44.

27. Cunningham JR. Scatter-air ratios. Phys Med Biol 1972;17:42–51.

28. Chui C, Mohan R, Fontanela D. Dose computation for asymmetric fields defined by independent jaws. Med Phys 1986;15:92.

29. Rosenberg I, Chu JC, Saxena V. Calculation of monitor units for a linear accelerator with asymmetric jaws. Med Phys 1995;22:55–61.

30. Slessinger ED, Gerber RG, Harms WB, et al. Independent collimator dosimetry for a dual photon energy linear accelerator. Int J Radiat Oncol Biol Phys 1993;27(3):681–687.

31. Palta JR, Ayyangar KM, Suntharalingam N. Dosimetric characteristics of a 6 MV photon beam from a linear accelerator with asymmetric collimator jaws. Int J Radiat Oncol Biol Phys 1988;14:383–387.

32. Khan F. Dosimetry of wedged fields with asymmetric collimation. Med Phys 1993;20:1447.

33. Klein EE, Harms WB, Low DA, et al. Clinical implementation of a commercial multileaf collimator: dosimetry, networking, simulation, and quality assurance. Int J Radiat Oncol Biol Phys 1995;33:1195–1208.

34. Palta JR, Yeung DK, Frouhar V. Dosimetric considerations for a multileaf collimator system. Med Phys 1996;23(7):1219–1224.

35. Jordan TJ, Williams PC. The design and performance characteristics of a multileaf collimator. Phys Med Biol 1994;39:231–251.

36. Das IJ, Desobry GE, McNeeley SW, et al. Beam characteristics of a retrofitted double-focused multileaf collimator. Med Phys 1998;25(9):1676–1684.

37. Boyer A, Biggs P, Galvin J, et al. AAPM report 72: basic applications of multileaf collimators, report of Task Group 50. Published for the American Association of Physicists in Medicine. Madison, WI: Medical Physics Publishing, 2001.

38. Hanson WF, Berkley LW. Calculative technique to correct for the change in linear accelerator beam energy at off-axis points. Med Phys 1980;7(2):147–150.

39. Hanson WF, Berkley LW. Off-axis beam quality change in linear accelerator x-ray beams. Med Phys 1980;7(2):145–146.

40. Otto K. Volumetric modulated arc therapy: IMRT in a single gantry arc. Med Phys 2008;35(1):310–317.

41. Bedford JL, Warrington AP. Commissioning of volumetric modulated arc therapy (VMAT). Int J Radiat Oncol Biol Phys 2009;73(2):537–545.

42. Rao M, Yang W, Chen F, et al. Comparison of Elekta VMAT with helical tomotherapy and fixed field IMRT: plan quality, delivery efficiency and accuracy. Med Phys 2010;37(3):1350–1359.

43. Ling CC, Zhang P, Archambault Y, et al. Commissioning and quality assurance of RapidArc radiotherapy delivery system. Int J Radiat Oncol Biol Phys 2008;72(2):575–581.

44. Yu CX. Intensity modulated arc therapy with dynamic multileaf collimation: an alternative to tomotherapy. Phys Med Biol 1995;40(9):1435–1449.

45. Yu CX, Tang G. Intensity-modulated arc therapy: principles, technologies and clinical implementation. Phys Med Biol 2011;56(5):R31–R54.

46. Ling CC, Archambault Y, Bocanek J, et al. Scylla and Charybdis: longer beam-on time or lesser conformality–the dilemma of tomotherapy. Int J Radiat Oncol Biol Phys 2009;75(1):8–9.

47. Mehta M, Hoban P, Mackie TR. Commissioning and quality assurance of RapidArc radiotherapy delivery system: in regard to Ling et al. (Int J Radiat Oncol Biol Phys 2008;72:575–581): absence of data does not constitute proof; the proof is in tasting the pudding. Int J Radiat Oncol Biol Phys 2009;75(1):4–6.

48. Kung JH, Chen GTY, Kuchnir FK. A monitor unit verification calculation in intensity modulated radiotherapy as a dosimetry quality assurance. Med Phys 2000;27(10):2226–2230.

49. Yang Y, Xing L, Li JG, et al. Independent dosimetric calculation with inclusion of head scatter and MLC transmission for IMRT. Med Phys 2003;30(11):2937–2947.

50. Purdy JA. Buildup/surface dose and exit dose measurements for 6-MV linear accelerator. Med Phys 1986;13:259.

51. Rustgi SN, Rodgers JE. Improvement in the buildup characteristics of a 10-MV photon beam with electron filters. Phys Med Biol 1985;30:587.

52. Gerbi BJ, Meigooni A, Khan FM. Dose buildup for obliquely incident photon beams. Med Phys 1987;14:393–399.

53. Klein EE, Purdy JA. Entrance and exit dose regions for Clinac-2100 C. Int J Radiat Oncol Biol Phys 1993;27:429–435.

54. Cunningham JR. Tissue inhomogeneity corrections in photon-beam treatment planning. In: C.G. Orton, ed. Progress in medical physics, vol 1. New York: Plenum Press, 1982:103–131.

55. Kornelsen RO, Young MEJ. Changes in the dose-profile of a 10 MV x-ray beam within and beyond low density material. Med Phys 1982;9:114–116.

56. Epp ER, Lougheed MN, McKay JW. Ionization build-up in upper respiratory air passages during teletherapy units with cobalt-60 irradiation. Br J Radiol 1958;31:361.

57. Epp ER, Boyer AL, Doppke KP. Underdosing of lesions resulting from lack of electronic equilibrium in upper respiratory air cavities irradiated by 10 MV x-ray beams. Int J Radiat Oncol Biol Phys1977;2:613.

58. Klein EE, Chin LM, Rice RK, et al. The influence of air cavities on interface doses for photon beams (abstract). Int J Radiat Oncol Biol Phys 1993;27:419.

59. Rice RK, Mijnheer BJ, Chin LM. Benchmark measurements for lung dose corrections for x-ray beams. Int J Radiat Oncol Biol Phys 1988;15:399–409.

60. Das IJ, Kase KR, Meigooni AS, et al. Validity of transition-zone dosimetry at high atomic number interfaces in megavoltage photon beams. Med Phys 1990;17(1):10–16.

61. Sibata CH, Mota HC, Hoggins PD, et al. Influence of hip prostheses on high energy photon dose distribution. Int J Radiat Oncol Biol Phys 1990;18:455–461.

62. Thatcher M. Perturbation of Cobalt 60 radiation doses by metal objects implanted during oral and maxillofacial surgery. J Oral Maxillofac Surg 1984;42:108–110.

63. Niroomand-Rad A, Razavi R, Thobejane S, et al. Radiation dose perturbation at tissue-titanium dental interfaces in head and neck cancer patients. Int J Radiat Oncol Biol Phys 1996;34(2):475–480.

64. Reft C, Alecu R, Das IJ, et al. Dosimetric considerations for patients with HIP prostheses undergoing pelvic irradiation. Report of the AAPM Radiation Therapy Committee Task Group 63. Med Phys2003;30(6):1162–1182.

65. Klein EE, Kuske RR. Changes in photon dosimetry due to breast prosthesis. Int J Radiat Oncol Biol Phys 1993;25(3):541–549.

66. Tatcher M. A method for varying effective angle of wedge filters. Radiology 1970;97:132.

67. Philips Medical Systems Division. Product Data 764. Eindhoven, The Netherlands: Philips Medical Systems Division, 1983.

68. Petti PL, Siddon RL. Effective wedge angles with a universal wedge. Phys Med Biol 1985;30(9):985–991.

69. McCullough EC, Gortney J, Blackwell CR. A depth dependence determination of the wedge transmission factor for 4-10 MV photon beams. Med Phys 1988;15:621–623.

70. Sewchand W, Khan FM, Williamson J. Variations in depth-dose data between open and wedge fields for 4-MV x rays. Radiology 1978;127:789–792.

71. Abrath FG, Purdy JA. Wedge design and dosimetry for 25-MV x rays. Radiology 1980;136:757–762.

72. Leavitt DD, Martin M, Moeller JH, et al. Dynamic wedge field techniques through computer-controlled collimator motion and dose delivery. Med Phys 1990;17:87–91.

73. Klein EE, Low DA, Meigooni AS, et al. Dosimetry and clinical implementation of dynamic wedge. Int J Radiat Oncol Biol Phys 1995;31:583–592.

74. Akazawa PF, Roach MI, Pickett B, et al. Three dimensional comparison of blocked arcs vs four and six field conformal treatment of the prostate. Radiother Oncol 1996;41:83–88.

75. Wu Q, Manning M, Schmidt-Ullrich R, et al. The potential for sparing of parotids and escalation of biologically effective dose with intensity-modulated radiation treatments of head and neck cancers: a treatment design study. Int J Radiat Oncol Biol Phys 2000;46(1):195–205.

76. Wu Q, Mohan R, Morris M, et al. Simultaneous integrated boost intensity-modulated radiotherapy for locally advanced head-and-neck squamous cell carcinomas. I: dosimetric results. Int J Radiat Oncol Biol Phys2003;56(2):573–585.

77. Verbakel WF, Senan S, Cuijpers JP, et al. Rapid delivery of stereotactic radiotherapy for peripheral lung tumors using volumetric intensity-modulated arcs. Radiother Oncol 2009;93(1):122–124.

78. Georg D, Knöös T, McClean B. Current status and future perspective of flattening filter free photon beams. Med Phys 2011;38(3):1280–1293.

79. Emami B, Lyman J, Brown A, et al. Tolerance of normal tissue to therapeutic irradiation. Int J Radiat Oncol Biol Phys 1991;21:109–122.

80. Marks LB, Ten Haken RK, Martel MK. Guest editor’s introduction to QUANTEC: a users guide. Int J Radiat Oncol Biol Phys 2010;76(3, Suppl 1):S1–S2.

81. Rosenow UF, Valentine ES, Davis LW. A technique for treating local breast cancer using a single set-up point and asymmetric collimation. Int J Radiat Oncol Biol Phys 1990;19:183–188.

82. Marshall M. Three-field isocentric breast irradiation using asymmetric jaws and a tilt board. Radiother Oncol 1993;28:228–232.

83. Klein EE, Taylor M, Michaletz-Lorenz M, et al. A mono-isocentric technique for breast and regional nodal therapy using dual asymmetric jaws. Int J Radiat Oncol Biol Phys 1994;28:753–760.

84. Foo ML, McCullough EC, Foote RL, et al. Doses to radiation sensitive organs and structures located outside the radiotherapeutic target volume for four treatment situations. Int J Radiat Oncol Biol Phys1993;27:403.

85. Sohn JW, Suh JH, Pohar S. A method for delivering accurate and uniform radiation dosages to the head and neck with asymmetric collimators and a single isocenter. Int J Radiat Oncol Biol Phys1995;32:809–814.

86. Takahaski S. Conformation radiotherapy-rotation techniques as applied to radiography and radiotherapy of cancer. Acta Radiol Suppl 1965;242:1–142.

87. Hounsell AR, Sharrock PJ, Moore CJ, et al. Computer-assisted generation of multileaf collimator settings for conformation therapy. Br J Radiol 1992;65:321–326.

88. Galvin JM, Smith AR, Lally B. Characterization of a multileaf collimator system. Int J Radiat Oncol Biol Phys 1993;25:181–192.

89. Boyer AL, Ochran TG, Nyerick CE, et al. Clinical dosimetry for implementation of a multileaf collimator. Med Phys 1992;19(5):1255–1261.

90. Powlis WD, Smith AR, Cheng E, et al. Initiation of multileaf collimator conformal radiation therapy. Int J Radiat Oncol Biol Phys 1993;25:171–179.

91. LoSasso T, Kutcher GJ. Multi-leaf collimation vs. Cerrobend blocks: analysis of geometric accuracy. Int J Radiat Oncol Biol Phys 1995;32:499–506.

92. Zhu Y, Boyer AL, Desorby GE. Dose distributions of x-ray fields as shaped with multileaf collimators. Phys Med Biol 1992;37:163–173.

93. LoSasso T, Chui CS, Kutcher GJ. The use of a multi-leaf collimator for conformal radiotherapy of carcinomas of the prostate and nasopharynx. Int J Radiat Oncol Biol Phys 1993;25:161–170.

94. Brahme A. Optimization of stationary and moving beam radiation therapy techniques. Radiother Oncol 1988;12:129–140.

95. Du MN, Yu CX, Symons M, et al. A multi-leaf collimator prescription preparation system for conventional radiotherapy. Int J Radiat Oncol Biol Phys 1995;32:513–520.

96. Powers WE, Kinzie JJ, Demidecki AJ, et al. A new system of field shaping for external-beam radiation therapy. Radiology 1973;108:407–411.

97. Leavitt DD, Gibbs FA Jr. Field shaping. In: Purdy JA, ed. Advances in radiation oncology physics: dosimetry, treatment planning, and brachytherapy. New York: American Institute of Physics, 1992:500–523.

98. Huen A, Findley DO, Skov DD. Attenuation in Lipowitz’s metal of x-rays produced at 2, 4, 10, and 18 MV and gamma rays from cobalt-60. Med Phys 1979;6:147–148.

99. Ellis F, Hall EJ, Oliver R. A compensator for variations in tissue thickness for high energy beams. Br J Radiol 1959;32:421–422.

100. Reinstein LE. New approaches to tissue compensation in radiation oncology. In: Purdy JA, ed. Advances in radiation oncology physics: dosimetry, treatment planning, and brachytherapy. New York: American Institute of Physics, 1992:535–572.

101. Chang SX, Cullip TJ, Deschesne KM, et al. Compensators: an alternative IMRT delivery technique. J Appl Clin Med Phys 2004;5(3):15–36.

102. Humphries SM, Boyd K, Cornish P, et al. Comparison of super stuff and paraffin wax bolus in radiation therapy of irregular surfaces. Med Dosimetry 1996;21(3):155–157.

103. Gerber RL, Marks JE, Purdy JA. The use of thermal plastics for immobilization of patients during radiotherapy. Int J Radiat Oncol Biol Phys 1982;8:1461.

104. Johnson MW, Griggs MA, Sharma SC. A comparison of surface doses for two immobilizing systems. Med Dosimetry 1995;20(3):191–194.

105. Higgins DM, Whitehurst P, Morgan AM. The effect of carbon fiber couch inserts on surface dose with beam size variation. Med Dosimetry 2001;26(3):251–254.

106. Keys R, Grigsby PW. Gapping fields on sloping surfaces. Int J Radiat Oncol Biol Phys 1990;18:1183–1190.

107. Johnson JM, Khan FM. Dosimetric effects of abutting extended source to surface distance electron fields with photon fields in the treatment of head and neck cancers. Int J Radiat Oncol Biol Phys1994;28:741–747.

108. Fraass BA, Tepper JE, Glatstein E, et al. Clinical use of a match line wedge for adjacent megavoltage radiation field matching. Int J Radiat Oncol Biol Phys 1983;9:209–216.

109. Christopherson D, Courlas GJ, Jette D. Field matching in radiotherapy. Med Phys 1984;3:369.

110. Siddon RL, Tonnesen GL, Svensson GK. Three-field techniques for breast treatment using a rotatable half-beam block. Int J Radiat Oncol Biol Phys 1981;7:1473–1477.

111. Williamson TJ. A technique for matching orthogonal megavoltage fields. Int J Radiat Oncol Biol Phys 1979;5:111.

112. Lim MLF. A study of four methods of junction change in the treatment of medulloblastoma. Am Assoc Med Dosim J 1985;10:17–24.

113. National Heart, Lung, and Blood Institute (NHLBI). What is a pacemaker? 2011. Available at: http://www.nhlbi.nih.gov/health/dci/Diseases/pace/pace_whatis.html

114. National Heart, Lung, and Blood Institute (NHLBI). What is an implantable cardioverter defibrillator? 2011. Available at: http://www.nhlbi.nih.gov/health/dci/Diseases/icd/icd_whatis.html

115. Hurkmans CW, Scheepers E, Springorum BGF, et al. Influence of radiotherapy on the latest generation of implantable cardioverter-defibrillators. Int J Radiat Oncol Biol Phys 2005;63(1):282.

116. Solan AN, Solan MJ, Bednarz G, et al. Treatment of patients with cardiac pacemakers and implantable cardioverter-defibrillators during radiotherapy. Int J Radiat Oncol Biol Phys 2004;59(3):897–904.

117. Sundar S, Symonds RP, Deehan C. Radiotherapy to patients with artificial cardiac pacemakers. Cancer Treat Rev 2005;31(6):474–486.

118. Marbach JR, Sontag MR, Van Dyk J, et al. Management of radiation oncology patients with implanted cardiac pacemakers: report of AAPM Task Group No. 34. Med Phys 1994;21(1):85–90.

119. Hudson F, Coulshed D, D’Souza E, et al. Effect of radiation therapy on the latest generation of pacemakers and implantable cardioverter defibrillators: a systematic review. J Med Imaging Radiat Oncol2010;54(1):53–61.

120. Stovall M, Blackwell CR, Cundiff J, et al. Fetal dose from radiotherapy with photon beams: report of AAPM Radiation Therapy Committee Task Group No. 36. Med Phys 1995;22:63–82.

121. Mutic S, Klein EE. A reduction in the AAPM TG-36 reported peripheral dose distributions with tertiary multileaf collimation. Int J Radiat Oncol Biol Phys 1999;44(4):947–953.

122. Francois P, Beurtheret C, Dutreix A. Calculation of the dose delivered to organs outside the radiation beams. Med Phys 1988;15(6):879–883.

123. van der Giessen PH. A simple and generally applicable method to estimate the peripheral dose in radiation teletherapy with high energy x-rays or gamma radiation. Int J Radiat Oncol Biol Phys1996;35(5):1059–1068.

124. Niroomand-Rad A, Cumberlin RL. Measured dose to ovaries and testes from Hodgkin’s fields and determination of genetically significant dose. Int J Radiat Oncol Biol Phys 1993;25(4):745–751.

125. Fraass BA, Kinsella TJ, Harrington ES, et al. Peripheral dose to the testes: the design and clinical use of a practical and effective gonadal shield. Int J Radiat Oncol Biol Phys 1985;11(3):609–616.