Total body irradiation (TBI) refers to a complex treatment modality that delivers a relatively uniform dose (± 10%) of radiation to the entire patient body.
TBI is often used in the treatment of various leukemias, lymphomas, and other hematological disorders. TBI is given as part of preparatory regimens for bone marrow transplantation in both adult and pediatric patients. These regimens typically employ high dose chemotherapy administered in conjunction with radiation. The purpose of TBI is twofold: (1) to contribute to the eradication of malignant cells or cells with genetic disorders and (2) to immunosuppress the patient sufficiently to prevent rejection of the donor bone marrow.
Although the preparatory regimen can be established with chemotherapy alone, there are some advantages associated with the addition of TBI: (1) the dose delivered to the patient’s body is relatively homogeneous and independent of the blood supply, (2) there is no sparing of sanctuary sites which are inaccessible to chemotherapy, and (3) the dose can be customized by shielding critical structures or by boosting areas of greater recurrence risk.
Equipment and Irradiation Techniques
Currently, most of the TBI procedures are performed with megavoltage photon beams from linear accelerators that are used for conventional radiation therapy treatments. Unconventional geometries are used to provide the desired large fields, typically 130 cm × 130 cm or greater.
Many different techniques have been reported for the effective irradiation of the whole body (Khan et al. 1980; Van Dyk et al. 1986). They can be divided into two groups: (1) those that utilize large field sizes that encompass the entire patient and (2) those that use multiple smaller adjacent fields to cover the patient’s body.
Multiple adjacent fields are used less frequently as they present two distinct problems. First, matching adjacent fields within the accepted dose uniformity is more difficult for large fields due to large divergence of the beam edges and the large patient volumes at the junction region. Second, the cells circulating throughout the body can potentially receive a reduced dose.
The large fields that encompass the entire patient’s body can be secured by positioning the patient at an extended distance, typically 3–4 m from the target. Treatments are usually delivered with a horizontal beam directed at a wall which serves as a primary barrier. These techniques are the easiest to employ and most well-established today.
The largest field size on a linear accelerator, 40 cm × 40 cm at the isocenter (1 m from the target), will project to a larger field size, e.g., 160 cm × 160 cm, using an extended source to surface distance (SSD), e.g., 4 m. Also by turning the collimator 45° the patient will lie along the square field diagonal. Positioning the patient within this field should be done carefully since the useful treatment field (90% uniformity) is substantially smaller than the projected light field. Film dosimetry at the time of TBI commissioning can be used to determine how far inside the light field the 90% decrement line lies.
Popular TBI techniques used today have the patient irradiated with parallel-opposed beam configurations: anterior-posterior (AP/PA) or left and right lateral fields (LT LAT/RT LAT).
The lateral opposed beam technique (Fig. 2b) is usually performed with the patient seated or lying down in the supine position. The ability of the patient to tolerate certain positions must be considered when choosing which technique to use. It is possible to have the patient’s legs in a semi-collapsed or fetal position, to decrease the field size required to cover the whole body. Generally, bilateral irradiation requires smaller field sizes than the anterior-posterior technique. Also, careful arm positioning can be used to shield the lungs; the arms should follow the body contour anteriorly in such a way that they will shadow the lungs but not the spinal canal.
Lateral opposed beams will usually produce larger inhomogeneities in dose compared to AP/PA treatments due to larger variation in body thickness along the beam direction. The dose to the head and neck area can be 20–30% higher than that at mid-abdomen. Missing tissue compensators for the head, neck, and legs can be designed to improve dose uniformity (Galvin et al. 1980; Khan et al. 1980; Van Dyk et al. 1986).
As treatment times can be long and patient weakness a common occurrence due to the combined toxicity of TBI and chemotherapy, one should also take into consideration patient comfort, stability, and reproducibility of the TBI setup.
Dose and Dose Prescription
Depending on the specific clinical situation there is a wide range in the prescribed dose-fractionation schemes in use today.
TBI was originally delivered as a single fraction treatment. The total dose was predominantly limited by fatal pulmonary toxicity from interstitial pneumonitis (IP). Since toxicity is influenced by the dose rate and total dose, fractionated and hyperfractionated TBI techniques were subsequently introduced to reduce pulmonary toxicity. High dose TBI with doses ranging from 2 to 14 Gy delivered over 1 to 9 fractions as well as low dose TBI with doses of 0.10–0.15 Gy per fraction delivered over 10–15 fractions have been used (Van Dyk et al. 1986).
The generally accepted dose prescription method for TBI, recommended by the American Association of Physicists in Medicine (AAPM), uses a single point specification (Van Dyk et al. 1986). The estimated uncertainties in dose delivery must be specified. The point typically used for the dose prescription is midplane at the level of the umbilicus, regardless of the delivery technique. However, many other points have been used (average midline depth of the head, neck, chest, abdomen, etc.). The manner in which the dose is prescribed must be explicitly stated in the prescription.
Dose limits to specific critical organs (e.g., lungs, kidneys) usually is included in the dose prescription. Depending on the clinical requirements the dose rate may be specified as well. Most clinical protocols require low dose rates (8–15 cGy/min) at the prescription point.
The International Commission of Radiation Units and Measurements (ICRU) (ICRU 1976) has suggested an overall accuracy in dose delivery of ±5%. However for large field geometries this is often difficult to achieve. If the prescribed dose is well below the onset of normal tissue toxicity, or if the dose to normal tissue can be locally limited (e.g., blocking) this accuracy level can be eased. For TBI treatments, the AAPM suggests (Van Dyk et al. 1986) the use of APARA principle (As Precise as Readily Achievable, technical and biological factors being taken into account). In practice, the generally accepted dose inhomogeneity is ±10% relative to the prescription dose. Nevertheless, the acceptable dose uncertainty should be a clinical decision based on several variables: the available beam energy, the irradiation technique, any treatment room constraints, type of disease, prior radiation treatment history, etc.
One of the many factors that limit the dose homogeneity in TBI is related to the variations in patient thickness. Consequently, irradiation techniques that minimize these variations are more preferable (e.g., AP/PA fields instead of bilateral fields). When treatment room constraints do call for the use of bilateral fields, dose variations can be decreased with the use of compensators or bolus material. Usually, the tissue compensators are one-dimensional compensators made of lead or copper, and are placed at the head of the machine. For a head and neck compensator, the position of the absorber is crucial; an underdose of the shoulders can occur if the compensator is shifted inferiorly or an overdose to the neck can occur if the compensator is shifted superiorly. The tissue-equivalent bolus material is placed directly on the skin, thereby increasing the dose to the skin and decreasing the midplane dose.
Dose homogeneity for parallel-opposed beams can be improved by increasing the beam energy. As a result, there is increased skin sparing and the possibility of underdosing the patient surface. A large spoiler screen placed as close to the patient as possible (typically at distance of no more than 30 cm) can be used to contaminate the incident beam with secondary electrons thereby raising the surface dose to at least 90% of the prescribed dose. A beam spoiler with a low atomic number such as acrylic or Lucite and a thickness of 1–2 cm will adequately modify the buildup curve.
To further improve dose homogeneity, larger treatment distances should be used if allowed by the room geometry. Ideally distances of at least 3 m from the target to the patient midplane should be used, although techniques that use less than this have been reported (Glasgow 1982).
Shielding Critical Structures
During the course of radiation treatment some critical organs may need to be shielded to either improve dose homogeneity (reduced lung density) or reduce the dose to those particular organs (e.g., kidneys).
The lung is very sensitive to radiation and is easily affected by the therapy regimen (dose, dose rate and fractionation), so knowledge of the lung dose as well as the dose rate during treatment is critical. The lung dose can be estimated from anthropomorphic phantoms measurements or by using computational methods. Without compensating for air density, the dose inhomogeneity in the lungs can exceed the prescribed dose by 10–24% (Van Dyk et al. 1986).
Lung dose can be lowered by using shielding blocks for part of the treatment. This dose should satisfy two criteria: (1) it should be sufficiently high to suppress the immune system and (2) it should be below the dose that causes the onset of radiation pneumonitis.
The shape of the blocks can be determined from radiographic films taken in both anterior and posterior positions. The lung blocks can be placed on the treatment unit head by using a block tray, or directly on the beam spoiler using heavy duty Velcro. Port films are commonly used to check the positioning of the blocks. Since immobilization of TBI patients is not always practiced (patients need to vomit with no advanced notice), patient movement can create a significant problem.
The use of lung blocks results in an overall reduction of the lung dose but does not decrease the dose variation throughout the organ itself. A lower and uniform dose to the lungs can be achieved by using lung compensators although the production of compensators can be a time consuming process.
Depending on the treated disease, doses to other critical structures (e.g., heart, kidneys, liver, and brain) may need to be limited.
If lung blocks are used, the chest wall under the blocks may be boosted with additional electron fields to avoid a decrease in dose to the marrow in the ribs (Van Dyk et al. 1986). The dose prescribed to the inner chest wall is typically 6.0 Gy, delivered over two fractions (Shank and Simpson 1982). The energy of the electron beam is selected to position the 90% isodose line at the lung–chest wall interface.
Due to leukemic relapse in the testes, an electron boost treatment may be added for the male patients. An enface electron beam of appropriate energy, to deliver up to 4.0 Gy, in one fraction to the posterior surface of the testes can be used (Shank and Simpson 1982). Similarly, the energy of the electron beam is selected to position the 90% isodose line at the posterior aspect of the testes.
It has been shown that it is best to design kidney blocks with the patient in the treatment position (Reiff et al. 1999). When compared to the supine position, patients in the upright position show a dramatic inferior shift of the kidneys with other obvious, but less predictable, changes. For patients who must be treated lying on their side, similar shifts in the kidneys size, shape, and position occur. For TBI treatment delivered in non-supine positions, kidney blocks should not be designed on the basis of supine abdominal CT scans.
The dosimetry of the large fields at extended treatment distances is significantly different than the conventional short distance treatments. TBI deviates from the standard radiation therapy techniques in multiple ways. In TBI the radiation field is larger than the irradiated volume, which is highly irregular in shape. Also, when dosimetric parameters are measured, a long length of the ionization chamber cable is irradiated in the large field geometry. Special attention should be given to ensure that the stem and cable effects are low and not impacting the measured data. The irradiated cable length should be kept as small as possible and the cable should be covered with buildup material to ensure electronic equilibrium within the cable.
Machine calibration, dosimetry, and monitor unit calculations have been extensively discussed in the literature (Van Dyk et al. 1986; Glasgow 1982; Podgorsak et al. 1985). The basic dosimetric parameters to be measured for TBI are the same as those for standard treatments, including absolute beam output calibration, percent depth dose (PDD), tissue-maximum ratio (TMR), tissue-phantom ratio (TPR), beam profiles, etc.
The minimum recommended phantom size for calibration purposes is 30 cm × 30 cm × 30 cm, although when available, larger phantoms should be used to ensure full scattering conditions (Van Dyk et al. 1986). Calibration is best to be performed under actual treatment conditions including the use of any treatment aids which will be used clinically such as shields, compensators, beam spoiler, etc.
Regardless of which dose ratio parameter will be used for TBI dosimetry, special attention should be given to its characteristics under these unusual treatment conditions. For conventional treatments, some dose ratio parameters (TMR, TPR) are considered independent of distance. The validity of this statement should be confirmed for the particular TBI geometry to be used. Also, dose ratio parameters requiring in air measurements (e.g., tissue-air ratio TAR) should not be used since they can be affected by backscatter from the treatment room wall. Furthermore, if percentage depth dose ratios are used, they should be measured for the particular treatment geometry since the use of Mayneord’s factors can be erroneous at these very extended distances. The simple extrapolation of dosimetric data from conventional field sizes and treatment distances is not acceptable.
After a particular technique has been commissioned for clinical use, the calculated doses should be confirmed by performing measurements in anthropomorphic phantoms as well as in vivo verification in patients. Thermoluminescent dosimeters (TLD’s) are the usual choice for monitoring the dose to the patient. Direct measurement of midline doses in patients can be performed at limited locations (mouth, between the legs or the feet, near the axilla for bilateral field irradiation). In most cases, the midline dose can be estimated from entrance and exit surface measurements. Surface measurements should be done with adequate build-up material to provide full electronic equilibrium. Corrections should be made to the exit dosimeter’s reading for the lack of full scatter at this position. Accurate placement of the exit dosimeter is also very important and port films can be used to confirm it.
TBI Implementation into Clinic
Any TBI protocol should be set up as a special treatment procedure following a careful plan of implementation, with attention to the nature of the patients, the limitations of the treatment room, as well as the available equipment.
First, a medical decision must be made regarding the total dose to be prescribed, the dose-fractionation scheme, the allowable dose inhomogeneity, the dose rate at the prescription point, and the dose to the critical structures.
Second, a treatment technique should be instituted. Frequently, conventional treatment facilities do not have the optimum extended SSD to achieve the large radiation fields required for TBI treatments. Hence, the available SSD may be partially responsible for the selected TBI technique.
At extended distances, the patient is typically set up against a primary barrier wall. To deliver the desired dose at the distance of the patient, the weekly and hourly workloads at the isocenter will increase so that special consideration should be given to the existing shielding characteristics of the primary barrier. Also, the composition of the primary barrier has to be taken into account since this can influence the dose the patient receives from scattered neutrons. Steel or lead slabs present in the primary barrier will result in an increased neutron dose equivalent delivered to the patient compared to a concrete barrier.
When new rooms are designed for TBI, ideally large SSDs can be achieved and special attention can be given to the room shielding. For example, the accelerator can be offset 1 m from the room center to allow a larger SSD to one wall.
A comprehensive quality assurance program should be set up to cover performance of the equipment used for treatment planning and dose delivery. This program should be extended to treatment aids as well: positioning equipment, blocks, compensators, block cutter, etc. Furthermore an in vivo dose measurement technique should be available when setting up a new TBI program. It is a good practice to verify the patient dose on the first treatment fraction.