31 Total Body Irradiation Conditioning Regimens in Stem Cell Transplantation
Kathryn E. Dusenbery and Bruce J. Gerbi
K. E. Dusenbery, MD
University of Minnesota Medical School, Department of Ra- diation Oncology, MMC436, 420 Delaware St. S.E., Minneapo- lis, MN 55455, USA
B. J. Gerbi, PhD
Associate Professor, Therapeutic Radiology – Radiation On- cology, University of Minnesota, Mayo Mail Code 494, 420 Delaware St SE, Minneapolis, MN 55455, USA
CONTENTS
31.1 Conditioning Regimen 786 31.2 Fractionation and Dose Rate 787 31.3 Sequence 788
31.4 Technical Aspects 788 31.5 Right and Left Lateral TBI 788
31.6 Simulation and Patient Measurements 788 31.6.1 Compensators for TBI 790
31.7 Compensator Design 791 31.7.1 Patient Treatment 793 31.7.2 Dose Verification 793 31.7.3 Dose Prescription 794 31.8 Anteroposterior TBI 794 31.8.1 Patient Treatment Technique 797 31.9 Normal Tissue Shielding 798 31.9.1 Lung Shielding 798
31.9.2 Kidney Shielding 799 31.10 Gonad Shielding 800 31.10.1 Thymus Shielding 800 31.10.2 TomoTherapy 800
31.10.2.1 Special Considerations for TBI in Young Children 801
31.10.2.2 Complications Following Preparation With TBI 802 References 802
immune (Thomas 1997) or genetic disorders are being offered transplantation (Iannone et al. 2003;
Peters et al. 2003). The rationale for SCT differs depending on the disease treated and the source of bone marrow cells. In both autologous and alloge- neic SCT for malignant diseases, the rationale for SCT is to allow chemotherapeutic dose escalation.
The SCT “rescues” the patient from what otherwise would be a lethal dose of chemotherapy.
In an allogeneic SCT, a healthy donor marrow regenerates in its place and the infused donor lymphocytes have a proven anti-tumor effect (graft versus leukemic effect) (Schleuning 2000;
Remberger et al. 2002; Zecca et al. 2002). Donor lymphocytes can also serve to supply absent enzyme for patients with inborn errors of metabo- lism. In autologous SCT, the infused marrow may be contaminated with malignant cells. Various methods have been used to purge the marrow of residual malignant cells (Freedman et al. 1998;
Colombat et al. 2000; Schouten et al. 2000; van Besien et al. 2003). As there is no anti-tumor (graft versus leukemia) effect, various cytokines are being tried in an effort to mimic the graft versus leukemia effect and improve the efficiency of autologous transplantation (Hawley et al. 1996;
Imamura et al. 1996; Klingemann 1996; Leshem et al. 2000). In the future, the autologous cells may be manipulated with genes that confer relative che- motherapeutic resistance (Wood and Prior 2001), thus allowing for additional post transplantation chemotherapy with little damage to the new bone marrow (Heslop et al. 1995).
The development of HLA (human lymphocyte antigen) and MLC (mixed lymphocyte culture) assays allowed physicians to determine whether potential marrow donors were “histocompatible”
(i.e., matched by HLA antigens and non-reactive to MLC cultures) and therefore less likely to develop graft-versus-host disease (GVHD) (Mahmoud et al. 1985; Parr et al. 1991; Petersdorf et al.
1998). Initially, transplants were only performed between HLA-matched related donors, but, with Since the first successful bone marrow transplanta-
tion was performed at the University of Minnesota
in 1960 (Gatti et al. 1968), bone marrow and stem
cell transplantation (SCT) has gained prominence
as a therapy for a variety of diseases as outlined in
Table 31.1. The majority of bone marrow transplants
are carried out in an effort to eradicate malignant
disease, but a growing number of patients with auto-
the advent of effective therapies directed at decreas- ing the probability and severity of GVHD, matched unrelated donor transplants have become more common. The National Marrow Donor Program (NMDP) types potential bone marrow and stem cell donors. More than 16,000 transplants have been performed using unrelated donors provided by the NMDP (McCullough et al. 1989; Karanes 2003;
Cornetta et al. 2005). With more than four million donors listed in the registry, over 70% of patients can now find an HLA-A, -B, -DR phenotypic match at the initial search.
In recent years, in addition to related and unre- lated bone marrow donor sources, additional sources of pluripotential stem cells have been inves- tigated, including the use of umbilical cord blood (Cornetta et al. 2005), a rich source of stem cells.
A bank of HLA typed umbilical cord blood harvests has also been established (Krishnamurti et al.
2003).
Another source of pluripotential stem cells are circulating blood stem cells. These cells can be har- vested through leukapheresis, frozen for later use, then thawed and reinfused. In addition to sparing the donor the discomfort of a bone marrow har- vest, these peripheral stem cell harvests usually result in a more prompt engraftment than occurs with bone marrow infusions, resulting in a shorter period of pancytopenia and thus less risk of infec- tion (Korbling et al. 1991; Steingrimsdottir et al.
2000).
31.1
Conditioning Regimen
Presumed desired endpoints of the pre-transplan- tation conditioning regimen are to eradicate the recipient’s native bone marrow, immune suppress the recipient sufficiently to avoid rejection of the donor transplant, and to do this with minimal toxic- ity to other tissues. In some situations, these three goals are attainable with a chemotherapy-only con- ditioning regimen; however, multiple variables need to be considered including the age of the patient, the underlying disease, the source of the donor marrow, and whether the donor marrow is manipulated (i.e., T depleted) (Fehr and Sykes 2004).
At the University of Minnesota, total body irradia- tion (TBI) is generally part of the conditioning regi- mens for the situations outlined in Table 31.2. These situations include unrelated marrow or cord blood donor transplants, certain underlying malignancies that are considered radiosensitive [acute lymphocytic leukemia (ALL), multiple myeloma], and for patients in whom a TBI-containing conditioning regimen has been shown to be superior to a chemotherapy-only conditioning regimen [acute myeloid leukemia (AML) in second remission] (Dusenbery et al. 1996).
There are theoretical advantages and disadvan- tages of a TBI-containing preparative regimen. Most often patients undergoing SCT have been exposed to multiple chemotherapeutic regimens; therefore, potential SCT recipients may be relatively chemo- therapy “resistant.” As most patients undergo- ing SCT have not been irradiated previously, their malignant cells may be more radiation sensitive than chemotherapy sensitive. Additionally, there are known sanctuary sites where chemotherapy does not penetrate well, such as the central nervous system (CNS) or testicles. There are no sanctuary sites for irradiation and, in certain situations such as relapsed ALL, a TBI-containing regimen may be especially beneficial. Lastly, chemotherapy which is usually given intravenously (busulfan may be given orally) needs to be metabolized and eliminated from the body. It is known that chemotherapy pharma- cokinetics differ among patients resulting in some areas of the body exposed to higher or lower con- centrations of drug. TBI requires no metabolism for clearance and all areas of the body receive the same dose of irradiation. The disadvantages of using a TBI regimen are the potential late side effects, such as sterility, cataracts, and growth retardation, as well as the potential neurological toxicity that may occur with irradiation.
Table 31.1. Diseases treated by stem cell transplantation Acute myeloid leukemia
Acute lymphoblastic leukemia Chronic myelogenous leukemia Chronic lymphocytic leukemia Myelodysplasia
Lymphoma Non-Hodgkin‘s Hodgkin‘s disease Multiple myeloma Aplastic anemia Idiopathic Fanconi
Paroxysmal nocturnal hemoglobinuria Congenital/immunodeficiency
SCIDS
Autoimmune disease Rheumatoid arthritis
Systemic lupus erythematosus Osteopetrosis
Leukencephalopathies Hurlers syndrome
Other inborn errors of metabolism
Sickle cell anemia/thalassemia
31.2
Fractionation and Dose Rate
Initial trials with TBI use a single fraction of up to 10 Gy. Subsequently, it was suggested that the thera- peutic ratio (i.e., increased leukemic cell kill and decreased toxicity of late responding tissues, such as lung, heart, spinal cord, kidney, and CNS) of TBI might be improved by either going to a very low dose rate (not practical since treatment times of up to 24 h would be required) or using a fractionated schema.
This was based on radiobiological data suggesting that leukemic cells (and their normal counterparts, the hematopoietic stem cells) were relatively radio- sensitive, with a narrow “shoulder” on the survival curve and with little capacity of sublethal radiation damage repair capacity (Hall 1994). However, late responding tissues are better able to repair sublethal damage and have a relatively “broad shoulder” on the dose–response curve. Formulations based on dose survival models have been proposed to evalu- ate the biological equivalence of various doses and fractionation schedules. Assumptions are based on the linear-quadratic model that takes into account the α and β (non-reparable and reparable) compo- nents of cell kill. The values for the α and β compo- nents of cell kill can be derived experimentally, but are not available for many human tissues. Extrapo- lating from animal data and cell cultures, it has been found that the ratio of α/ β is a useful indicator of the effect of fractionation on cell damage. Tissues with a high α/ β include the gastrointestinal tract,
skin, and bone marrow cells (“early” responding);
whereas, tissues with a low α/ β include spinal cord, kidney, brain, and lung (“late” responding) (Hall 1994).
Although not completely applicable in the setting of TBI schemes, the biological effective dose (BED) of one TBI regimen can be compared with another regimen if several estimates are taken into consider- ation. The units are arbitrary but allow one to com- pare the theoretical effects of different TBI regimens on different tissues.
BED = n × d (1+d/(α/β))
where n=number of fractions, d=dose per fraction (Gy/fraction) and α/β=estimate (10 for early tissues, 3 for late tissues).
In engineering an optimal TBI regimen, the goal is to cause minimal damage to late responding tis- sues, while having a high probability of damaging the early responding bone marrow cells (and malig- nant cells). Regimens can be evaluated for their potential effect on late responding tissues or on early responding tissues. If evaluating for late effects of different TBI regimens:
Assume an α/β of 3 BED 750 cGy in 1 fraction 26 750 cGy in 3 fractions 14 1200 cGy in 6 fractions 20 1320 cGy in 8 fractions 20 1395 cGy in 12 fractions 19
One can see that fractionation would be expected to spare late effects. One could go to a higher total dose, without increasing the probability of late effects. If considering the early effects of different TBI regimens:
Assume an α/β of 10 BED 750 cGy in 1 fraction 13 750 cGy in 3 fractions 9 1200 cGy in 6 fractions 14 1320 cGy in 8 fractions 15 1375 cGy in 12 fractions 16
It becomes apparent that fractionation spares early effects. Since bone marrow ablation is desired, the total dose must be increased to 1200 cGy or more to get the same myeloablative effects as a single frac- tion of 750 cGy.
These theoretical equations are supported by data from reports of bone marrow transplantation
Table 31.2. University of Minnesota protocols utilizing total body irradiation or total lymphoid irradiation in the condi- tioning regimen (CR)
1320 cGy in eight fractions over 4 days:
– Acute lymphoblastic leukemia in second or subsequent CR – Acute lymphoblastic leukemia high risk in first CR – Chronic myelogenous leukemia
– Acute myeloid leukemia (some children receive chemo- therapy only conditioning)
– Myelodysplastic syndrome
– Non-Hodgkin‘s lymphoma (depending on dose of prior irradiation)
– Unrelated donor transplantation – Cord blood transplantation Other regimens
– Fanconi anemia: 450 cGy in single fraction, ±thymus block – Non-myeloablative transplants: 200 cGy in a single fraction – Sickle cell anemia or thalassemia: total lymphoid irradiation,
block gonads
– Osteopetrosis: TLI to include spleen, liver and mesenteric
lymph nodes
regimens. For example, the risk of cataract for- mation (a late responding tissue) is substantially higher when the TBI is given in a single fraction of TBI than when it is fractionated (Aristei et al.
2002).
Although no randomized clinical trials exist, the majority of retrospective reviews looking at the rate of interstitial pneumonitis after single fraction TBI compared with multiple fraction TBI strongly suggest an advantage for a fractionated TBI schedule (Shank et al. 1983; Cardozo et al.
1985; Kim et al. 1985; Molls et al. 1986; Standke 1989; Valls et al. 1989; Carlson et al. 1994). One must keep in mind the fact that these trials are non-randomized and usually compare recently transplanted patients on fractionated schemas with former single fraction patients. Numerous variables are potentially implicated in the devel- opment of interstitial pneumonitis, including other TBI variables such as dose rate, use of lung shielding, and timing of the TBI (before or after chemotherapy) (Molls et al. 1986).
Initially single fraction TBI was delivered with cobalt units at extended distances. The dose rate achievable was only 5–10 cGy/min and it required several hours to deliver the dose. When using fractionated TBI, it is not clear that the dose rate needs to be this low, although the risk of pneumo- nitis may be higher with higher dose rate (Kim et al. 1985; Carruthers and Wallington 2004).
However, most institutions have kept the dose rate low (under 10 cGy/min), since a fraction of 200 cGy can be delivered in a reasonable amount of time (20 min in this case).
31.3 Sequence
TBI can either precede or follow the chemotherapy portion of the conditioning regimen. An advantage to delivering the TBI first is that, with the appro- priate use of antiemetics, it can be given as an out- patient treatment and thus reduce inpatient costs.
Following completion of TBI, patients are then hos- pitalized for the chemotherapy portion of the condi- tioning regimen. Clinical data are lacking, however, on whether TBI is less toxic or more effective when given before chemotherapy, although it is theoreti- cally possible that the variety of cytokines released during chemotherapy may influence the incidence of pneumonitis.
31.4
Technical Aspects
Numerous techniques for irradiation of the entire body are described in the literature. At the Univer- sity of Minnesota, two general TBI techniques are currently in use, with modifications of the tech- niques for certain situations. The vast majority of our patients are treated using the first technique, which involves right and left lateral fields with the patient semi-recumbent at an extended distance on a specially designed couch. The second tech- nique is an anterior–posterior treatment technique patterned after that developed at Memorial Sloan Kettering (Shank et al. 1983). For the latter, adult patients are treated in a standing position with ante- rior and posterior beams, while younger (smaller) patients are treated in a reclined position if they can fit within the available field size at the floor of the treatment room. The goal of both of these techniques is to deliver a uniform dose to the entire body within ±10% of the dose at the prescription point. This second technique has the advantage that certain organs such as the thymus or testicles can be blocked.
31.5
Right and Left Lateral TBI
This technique uses lateral photon beams with the patient in a semi-recumbent position, as described by Khan et al. (1980). The treatment is delivered at a source to patient midline distance of 410 cm, which produces a field approximately 120 cm wide at the 95% isodose line. Aluminum compensators are used to produce a uniform dose through all body regions to within +10% of the dose specified at the umbilicus.
31.6
Simulation and Patient Measurements
Pretreatment measurements for TBI are performed
in the simulator room to accurately reproduce the
treatment position within the treatment room and to
calculate the size and thickness of the compensating
filters. The simulation procedure consists of three
steps. During the first step, an anterior chest film is
taken to determine the amount of lung traversed by
the treatment fields. The radiograph is taken with the patient seated on the simulator couch with his or her back resting against the film cassette holder.
The source–skin distance (SSD) to the patient’s chest is measured and because the source–film distance (SFD) is known, the magnification factor for the size of the lungs can be determined. Using this informa- tion, the thickness of the lung inhomogeneity can be computed. This information is used later when the need for lung compensation is evaluated.
The second step is to establish the patient treat- ment position so that the entirety of the body fits within the 95% isodose line. An overhead projector is used to cast an isodose pattern that represents the uniformity of the actual treatment field (Fig. 31.1).
The head and back of the patient are positioned within the 98% line, while the toes of the feet are within the 95% isodose line of the radiation field.
Once this position has been established, setup mea- surements are recorded on the form, as illustrated in Figure 31.2.
To describe the position of the lower extremities, measurements are made from our reference point [the anterosuperior point of the iliac spine (ASIS)] to the knee and to the back of the heel. The length of the feet is also recorded. The distance from the sternal notch to the top of the knee or patella is documented to describe how compressed the patient is within the field. Finally, with the arms in the treatment position,
Fig. 31.1. For the bilateral technique, the patient is positioned within the homogeneous portion of the beam. The simulation is performed in the simulator room and an overhead projector is used to produce a representation of the treatment fi eld
Fig. 31.2. The form used to document the patient treatment position when the bilateral body technique is used
the distance from the middle of the shoulder to the top of the head is measured. The latter information is used to scale the size of the head compensator. The chin extension, measured from the sternal notch to the point of the chin, is also recorded. Additionally, the need for positional devices, such as pillows for the back of the head, foam sponges underneath the hips, or sandbags under the feet, is recorded on the form. For future reference, the names of the indi- viduals who made and checked the measurements are recorded.
The third step is to measure the right–left lateral
thickness of the patient at certain anatomical loca-
tions. These measurements are recorded on the form
shown in Figure 31.3 and constitute the basic data
needed to determine the thickness of the compensa-
tors at these locations. The key measurement is the
width of the patient at the umbilicus since this is the
location where the dose is prescribed. Values of lat-
eral thickness are also measured at the head, neck,
shoulders, mid-mediastinum, pelvis, knees, and
ankles. The mid-mediastinal thickness is measured
midway between the sternal notch and the xiphoid
and includes the thickness of the arms.
simulation. The form shown in Figure 31.3 also serves as the calculation sheet for the determi- nation of compensator thickness at the different locations. The compensators are usually designed in three pieces: one for the lower extremities, one for the head and neck region, and one for the lungs.
In most cases, a lung compensator is not required since the effective thickness at the mid-mediasti- num is usually greater than that at the umbilicus.
The arms are deliberately positioned in line with the lungs and act to increase the total thickness in this region.
The first step in designing tissue compensators is to determine the tissue deficit (TD), the difference in tissue-equivalent thickness between the prescrip- tion point (which in our case is the umbilicus) and the other locations. The following equation is used to calculate tissue deficit:
TD = L
ref− + − L ( 1 ρ
lung) L
lung(1)
where L
refis the lateral separation at the umbili- cus, L is the lateral separation at that particular anatomical location, L
lungis the separation of the lung determined from the anterior radiograph, and ρ
lungis the density of the lung. For ρ
lung, a value of 0.25 g/cm
3is used as the average lung density for healthy lung tissue (Van Dyk et al. 1982). Equa- tion 1 is used only for the mid-mediastinal location where lung tissue is present. At all other locations, L
lungis zero and the tissue deficit can be obtained using Eq. 2:
TD = L
ref− L (2)
L
lungis determined using the anterior chest radio- graph that was taken in the simulator. The lung thickness is determined at a point midway between the sternal notch and the most superior aspect of the domes of the diaphragms as seen on the radiograph.
As shown in Figure 31.4, two lateral measurements are made for both the right and left lobes of the lung:
the first measurement, represented by L
Rt1and L
Lt1, extends from the most lateral aspect to the most medial portion of the lung. The second measure- ment, represented by L
Rt2and L
Lt2, spans from the most lateral extent of the lung to the mediastinum.
The lung thickness is calculated using the following equation:
L L L L l SSD SFD
lung
SFD
rt rt lt lt
= ( + + + ) × ⎛ ⎝ ⎜⎜⎜ × + ⎞
⎠ ⎟⎟⎟
1 2 1 2
2
1
2 ( )
(3)
Fig. 31.3. The form used to record the values of right–left lateral thickness of the patient. This information is used to calculate the tissue defi cits that exist at various body locations versus the umbilicus thickness. Values of compensator thickness are subsequently calculated from these data. The fi nal column pro- vides a location to document the percentage of the prescribed dose delivered to the midline of the indicated regions
Once the lateral separations are recorded, the point where the lower extremities compensator is to start must be determined. Since the dose is pre- scribed for the midline thickness at the umbilicus, and the pelvis is usually of greater thickness, the compensator must be started at some point below the pelvis. The location where the compensator is to begin is that point on the legs that has the same sep- aration as the umbilicus. As final documentation, a photograph is taken with the patient in the treat- ment position with respect to the radiation field.
31.6.1
Compensators for TBI
Compensator thickness determination is calcu-
lated from measurements taken at the time of
where SSD is the source–skin distance and SFD is the source–film distance measured when the anterior chest radiograph was obtained during the first step of the simulation. In Eq. 3, the terms in the second parentheses serve to de-magnify the lung dimen- sions measured on the chest radiograph to life-size at the midplane of the patient.
The compensator thickness, L
c, is determined using the following equation:
L TD K
c
comp eff comp
= × × − ⎛ ×
⎝ ⎜⎜⎜
⎜
⎞
⎠ ⎟⎟⎟
⎟⎟
1 2
τ
ρ µ
τ ρ
ln (4)
In this expression, τ is the thickness ratio (Khan et al. 1970; Kirby et al. 1988). For beam energies from cobalt 60 (Co-60) to 10 MV, a value for τ of 0.70 is a good approximation for compensator distances greater than 20 cm from the surface (Khan et al.
1980). The density of the compensating material is p
comp(aluminum in this case), K is the off axis cor- rection factor that accounts for both the decreases in beam intensity away from the central axis and effec- tive scattering field size for the various locations, and µ
effis the broad-beam linear attenuation coefficient in tissue for this beam energy. The effective field size at various locations can be determined by Clarkson integration. The data in Table 31.3 were calculated in this manner for a Rando phantom and represents the equivalent scattering field at various locations of the body as a function of beam energy for an average adult (Kirby et al. 1988). This data can be used to determine accurate values for K in Eq. 4.
An alternate method to determine compensator thickness for TBI has been suggested by the Ameri- can Association of Physicists in Medicine (AAPM) (1986). Comparisons made between the system of
compensator thickness determination described above and that suggested by the AAPM (1986) show that the two systems produce compensators the thickness of which varies by less than 1 mm of alu- minum. Finally, the percentage difference in dose between the prescription point and other locations of the body are calculated and recorded in the last column of the form.
31.7
Compensator Design
The compensators are designed to be located at a distance of 72 cm from the virtual source of the accelerator so that appropriate compensation can be provided without the devices becoming excessively large and difficult to handle. Thus, the measurements recorded on the setup sheet in Figure 31.2 must be de-magnified from the treatment distance to the location of their use. Using the information supplied in Figures 31.2–31.4, the size of the compensators is determined in the following manner. For the lower extremities compensator, the base length for section A in Figure 31.5 is obtained by taking the ASIS to knee distance, subtracting the distance below the ASIS that the compensator is to start, and multiplying that dimension by the lateral magnification factor:
Lateral magnification factor = Source-Compensator tray distance
Source-axis distance- L
ref2
⎛⎛
⎝ ⎜⎜⎜ ⎞
⎠ ⎟⎟⎟
(5) Section B of the lower extremities compensator is the de-magnified distance from the knee to the back of the heels, while section C is the de-magnified length of the feet plus an additional 2 cm to ensure adequate coverage of the feet. Section D is simply an additional 2 cm of aluminum that is needed to clamp the compensator to the compensator tray during the actual treatment. The thickness of the compensator is obtained from the compensator thickness column as shown on Figure 31.3 for the corresponding ana- tomical location.
The head and neck compensator is designed in much the same manner. The base length of section E of this compensator is the de-magnified distance from the middle of the shoulder to the top of the head. Section F is an additional 1 cm of material to ensure adequate coverage, while section G is pro- vided for clamping. The thicknesses of the compen-
Table 31.3. The equivalent square field size for various ana- tomical locations. Determined at midline in a Rando phantom.
The lateral dimensions of the phantom at these locations is also listed in addition to the equivalent lengths required to obtain the calculated field sizes by equivalent square calcula- tion. Reproduced from Khan et al. (1970)
Photon energy Side of equivalent square (cm)
Head Neck Chest Umbilicus Hips
Co-60 17 22 31 30 28
4 MV 16 20 30 29 23
6 MV 18 23 29 27 26
10 MV 17 22 33 27 26
18 MV >18 >18 >18 >18 >18
Mean 17 22 31 28 26
Lateral dimension 15 12–16 31 27 31
Equivalent length 19 30 31 29 20
sator are again obtained from Figure 31.3 for the head and neck regions.
The length of the lung compensator is obtained by de-magnifying H
compfrom the chest radiograph (Fig. 31.4) to life size using (1/2)(SSD–SFD)/SFD, then again de-magnifying these dimensions to the treat- ment position of the compensators using Eq. 5. The compensator thickness is obtained from Figure 31.3.
The width of the compensators (the dimension of the compensator that is not shown in Fig. 31.5) is typically 11 cm for the lower extremities compensa- tor, 6.5 cm for the head and neck compensator, and 7.5 cm for the lung compensator.
Figure 31.6 shows the compensator design form that is sent to the machine shop for fabrication. Fig- ure 31.7a shows the finished aluminum compensa- tors and Figure 31.7b illustrates the compensators in use and how they are attached to the tray using specially designed clamps. The lung compensator, when required, is attached to the tray with double- sided tape.
Machine calibration and treatment calculation.
The linear accelerator is calibrated according to the protocol outlined by the AAPM (Almond et al.
1999). To determine the total number of monitor units (MUs) for TBI, the calculation is done as an isocentric treatment at an extended distance. Equa- tion 6 is used in the determination as follows:
MU = TD STF
k S r S r TMR d,r f
c o p o e
f
×
× × × × ⎛ ′
⎝ ⎜⎜⎜ ⎞
⎠ ⎟
( ) ( ) ( ) ⎟⎟⎟
⎛
⎝
⎜⎜⎜ ⎜⎜⎜
⎜⎜⎜ ⎜⎜
⎞
⎠
⎟⎟⎟ ⎟⎟⎟
⎟⎟⎟ ⎟⎟⎟
2