Development of a Translating Bed for Total Body Irradiation

Medical Dosime~y, Vol. 13, pp. 195-199 Printed in the U.S.A. All rights reserved.

DEVELOPMENT

0739-021 l/88 $3.00 + .oo Copyright 0 1988 American Association of Medical Dosimetrists

OF A TRANSLATING BED FOR TOTAL BODY IRRADIATION

W. LOCUS, L. JOHNSON, and E. SCHARTNER S. CONNORS,J. SCRIMGER, Dept. of Medical Physics, Cross Cancer Institute, Edmonton, Canada Abstract-Total

body irradiation is used to prepare a patient for bone marrow transplantation. Traditional techniques often sacrifice dose uniformity for patient comfort and ease of treatment. A method has been developed using a translational bed under a Cobalt 60 photon beam. The bed and controller were designed and built on site. A bolused patient lying in the bed is moved at constant speed through the beam. Using this technique, dose homogeneity is optimized by the use of bolus, extended source-skin distance, adequate field size and use of anterior/posterior fields. The dose rate represents a compromise between a value high enough to keep treatment times tolerable by the patient and one that is sufficiently low to avoid treatment complications. The value of 50 cGy/min which was used meets these requirements. Extensive phantom measurements have shown that the dose homogeneity can be obtained to within an acceptable limit of f5%.

variable speed d.c. motor which is coupled to a precision lead screw which runs the length of the frame. The turning lead screw drives a connecting nut which is attached to the bed top, allowing the latter to be moved along the entire length of the screw. The plastic frame shown in the diagram rests on the bed top and is secured into position with four locating pins situated at the corners of the top. The patient is placed into the frame, which can be adjusted to a suitable length with the divider shown. Bolus is built up around the patient. We have used rice bags as a bolus, and have been able to achieve good homogeneity and have eliminated irregular surface contours presented to the beam. For the PA field complete bolus around the head cannot be achieved due to restrictions to breathing, and the latter problem must be allieviated with the use of a head rest which raises the head away from the base, and a battery operated fan which blows fresh air through a vent in the frame. The compromise in bolus distribution in this region does not significantly detract from dose uniformity, as can be seen in the results.

INTRODUCTION

Total body irradiation (TBI) is often used in combination with chemotherapy prior to bone marrow transplantation for the treatment of acute leukemias. This treatment regimen prepares the patient for autologous, allogeneic or syngeneic bone marrow transplantation first by destroying the leukemic stem cells, immunocytes and bone marrow cells, and secondly by suppressing the immune reaction to the transplanted tissue. The radiotherapy dose should be delivered as homogeneously and precisely as possible to provide uniform cell killing and to avoid overdosing critical organs such as the lung. A variety of TBI procedures have been documented in the literature’ but often compromise on dose uniformity (7%-20%) and require either complex geometries or major modifications to the treatment machine. Choice of a particular method is limited by the facilities available and the compatibility of the chosen procedure with patients and staff. A method offering optimal dose homogeneity and control over dose delivery, while minimizing mechanical modifications to the treatment machine and at the same time providing a tolerable treatment time for the patient is presented here. It involves the translation of a bolused patient through a megavoltage photon beam.

B. Bed control system The bed system described above can be controlled both inside the treatment room, to set up the treatment, and outside the room. When controlled from inside the room the selection switch shown in Fig. 2 is set to the set-up position, and bed position is set with the hand held pendant shown in Fig. 2 (inset). Under these conditions, the bed can be moved back and forth to any desired position. The bed position is monitored and displayed at the control console (Fig. 2) through a pulse encoder incorporated in the drive motor. A programmable microprocessor allows start and stop positions of the bed to be set, and the desired drive speed can be selected at the speed potentiometer shown.

METHOD A. Bed construction The general layout of the bed is shown in Fig. 1. A frame with dimensions of 65 cm X 2 10 cm supports the drive system. This frame can be uniquely and reproducibly locked into a fixed position in the treatment room with locking bolts, one of which can be seen in the diagram. Preset holes in the floor accept these bolts. The drive system consists of a 4 HP 195

Medical Dosimetry

196

Volume 13, Number 4, 1988

Adjustable Divide Removable Sides

Cassette Slot

Fig. 3. Moving

Motor

Battery for Fan

Fig. 1. Diagram

of moving bed.

Bed Treatment

Technique.

of the patient, PI, and is entered into the microprocessor. The stop position, P2, is set such that the entire length of the patient will be driven through the field and 10 cm beyond. Thus P2 will have a value: P2 = (PI + 0.65 + 0.2 + h) meters

C. Method of operation The treatment technique used with the bed system described is illustrated in Fig. 3. The patient is placed into the frame, and is fully bolused with rice bags. The bed is then driven, with the hand control, until the head of the patient is 10 cm outside the edge of the 65 cm X 65 cm field, (as measured at the patient mid-plane). This constitutes the start position

With PI and P2 set, and the bed speed selected according to the required dose as discussed below, the beam is turned on and the bed motion started. The patient will then be driven through the beam at the pre-set speed, and the motion will stop and the beam shut off when the bed reaches P2. The patient is then turned over, re-bolused, and driven through the beam from P2 to P 1. Limit microswitches prevent the bed from driving beyond its mechanical range, and a motion stall sensor is interlocked to the treatment unit and causes the source to be retracted in the event of movement failure or obstruction. TREATMENT

PLANNING

A. Beam profile Measurement of beam flatness profiles eliminated concerns of the non-uniformity of the photon fluence and variations in dose rate in the treatment field. Transverse and longitudinal profiles were taken at depth and at extended distances in large tissueequivalent phantoms. It was found that the transverse and longitudinal profiles matched. Although a 15% drop in dose rate was evident at the outer edges of the 65 cm X 65 cm field, the profile showed a variation of less than f2.5% over the central 45 cm X 45 cm field (Fig. 4). The fall off at the outer edges of the beam will have little effect on dose homogeneity if the patient is started and stopped well outside the beam edge.

Fig. 2. Remote

control console of moving the local hand control.

bed. Inset shows

B. Dosimetry If the dose at midplane to the bolused patient can be determined accurately, then the time during which a patient must stay in the beam in order to receive a specified dose can be calculated. Given that any particular point must remain for a time which is determined by the ratio of the prescribed dose to the dose rate at midplane, and traverse a field of 65 cm, a

Total body irradiation 0 S. CONNORS et al. BEAMPROFILEFORCOBALT60 110IW-

Fig. 4. Beam Profile Graph: Measurements were performed in 1) polystyrene phantom (longitudinal scan) and 2) in a large water tank (transverse scan) at 0.5 cm depth and 150 cm SSD. A Capintec PR06C ionization chamber was used for the longitudinal measurements and a p-type diode for the transverse scan.

of the appropriate bed speed to deliver a specified dose is straightforward.

calculation

PRESCRIBED DOSE

Accurate determination of the dose rate at midplane is critical. In-air calibrations should be avoided, including the use of Tissue Air Ratios (TAR) because of the problem of floor and wall scatter. There is some

197

indication in the literature that the use of extreme treatment distances results in distance-dependent changes in Tissue Phantom Ratios (TPR) and Tissue Maximum Ratios (TMR).’ These parameters must therefore be measured for the user’s geometry, to enable one to make routine calculations of the dose at midplane. A full discussion of dosimetry problems can be found in the literature.2 For our purposes, an ion chamber measurement at midplane in a large polystyrene phantom surrounded by bolus was performed on the bed at the required geometry. The dosimetry protocol of TG-2 1 was applied. The measurement was within 2% of a calculated value using TAR values from the literature.’ Measurements of TPR are being undertaken for succeeding patients. C. Inhomogeneity corrections It has long been recognized that inhomogeneities such as lung and bone present special difficulties for treatment planning. The treatment planning system utilized in this study consists of a locally developed system known as ATP (Alberta Treatment Planning). This system uses an “equivalent TAR algorithm” for correcting doses to these organs. Initial treatment plans demonstrated that lung doses were unacceptably high (+20%) in some regions. Pulmonary toxicity has been shown to be a problem for TBI patients because of the high prescription dose to the target volume (bone marrow), and thus lung shielding is necessary. Lung shields of 4 mm lead were cut to the

Fig. 5. Isodose distribution of an adolescent child using Alberta Treatment Planning System. Two parallel opposed beams are used. Comparison of measured percentage depth dose curves in the treatment geometry with calculated data taken from the treatment planning system indicated a maximum of a 3% discrepancy. These distributions compared favorably to Rando phantom studies.

198

Medical Dosimetry

Fig. 6. Isodose distribution

Volume 13, Number 4, 1988

showing transverse slice with lung shields.

shape of the patient’s lung based on simulator fluo-

roscopy, and placed directly on the skin. Transmission measurements for 4 mm of lead were performed in phantom at extended distances at the required geometry. Attenuation was found to be 19.5% at 8.5 cm polystyrene depth (midplane for the first patient). Measurements on Rando phantom using these lung shields indicated a drop in lung dose to 85% of the delivered dose in agreement with the treatment plan (see Fig. 5). Bone inhomogeneity corrections are not generally as well described for treatment planning. While corrections for the change in photon fluence between bone/tissue interfaces are routinely handled by treatment planning computers, corrections for change in the photon spectrum at the depth of bony structure and subsequent absorption by the bone can be difficult to calculate since it requires detailed knowledge of the photon spectrum at depth. The issue is further complicated by the complex anatomical relationship of bone marrow and blood cells to the bone. This study did not correct for all these factors, other than applying an equivalent TAR correction from the treatment planning system. The dose to the bone marrow is not known precisely and we have assumed it is similar to the dose delivered to surrounding tissue. The problem is treated in some detail by reference 1. D. Alderson Rando phantom evaluation Thermoluminescent dosimeters were placed midplane in a Rando phantom and the phantom was given an AP/PA treatment with rice bolusing. The results are displayed in Fig. 7 and indicate a dose uniformity of f5% except to the lung shielded area.

FIRST CASE

HISTORY

A 10 year old male was our first patient to be treated, prior to his receiving an autologous bone marrow transplant. A fractionation scheme of 2 Gy per fraction, two fractions daily, to a total of 12 Gy was adopted. The bed speed was set at approximately 6 mm/set to achieve this dose and total treatment time was around 2 minutes per field. The patient tolerated the treatments well, although anti-emetics were used in the later stages of the treatment to prevent possible disruption of the treatment and to make the patient more comfortable during treatment. Staff adapted to the technique quickly, and by the last (third) day, the treatment was completed within 30 minutes. Figure 8 shows a TLD skin dose evaluation

TLD RAND0 EVALUATION Mid Plane Doses Lead Lung Shields,,-

,

APlPA Field

Fig. 7. TLD measurements on Rando Moving Bed Technique.

Phantom

using

Total body irradiation 0 S. CONNORS et al

199

TLD SKIN DOSE EVALLiAflaN

200 cGy/fraction

for a total of 1200 cGy

Fig. 8. TLD measurements

of the patient for the entire treatment. Note that the groin received a dose of 109%, outside our specification of 100% 6%. The TLD was placed on the inner thigh in the crotch area and well bolused. It is thought that the elevated dose is caused by increased scattered radiation due to air gaps or patient geometry. On succeeding patients, a flap of Superflab was placed on the TLD, and this has reduced the TLD readings to 100% as expected. CONCLUSIONS A moving bed technique under a megavoltage photon beam can be used to deliver a homogeneous

2 fractions/day

on 10 yr old male.

whole body dose within +5% without the necessity for modification of the treatment unit. The use of lung shields on the patient’s skin successfully decreases the lung dose and avoids the complications of pulmonary toxicity. REFERENCES 1. Van Dyk, J.; Galvin, J.M.; Glasgow, G.P.; Podgorsak, E.B. The physical aspects of total and half body photon irradiation. AAPM Report No. 17, New York, American Institute of Physics, 1986. 2. Quast, U. Physical treatment planning of total-body irradiation: Patient translation and beam zone method. Medical Physits, 12:567-574; 1985.