| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Correspondence |
North Western Medical Physics, Christie Hospital NHS Trust, Wilmslow Road, Manchester, M20 4BX , UK
(The Editors do not hold themselves responsible for opinions expressed by correspondents)
The Editor—Sir,
Lillicrap et al [1] have raised questions about the influence of leakage radiation from linear accelerators used for intensity modulated radiotherapy (IMRT). Specifically, they ask whether this new method of treatment will require a tightening of the leakage requirements and consequent revision of the International Electrotechnical Commission (IEC) standards [2]. Clinical implementation of IMRT will require consideration of many aspects of machine performance, including those of the acceptable leakage. Leakage in the patient plane is of particular importance, but leakage in other directions must be taken into account in the design of the secondary barriers in the treatment room [3]. We will briefly address both issues in this letter.
Delivery of a specified dose to the isocentre from a modulated field, delivered by either dynamic IMRT or the step and shoot method of IMRT, will, in general, require the accelerator to be energized for longer (hence more monitor units are needed) compared with delivering the same dose from an unmodulated field. It therefore follows that doses due to leakage radiation will be increased, although the spatial distribution and the magnitude of these doses will depend on many interrelated factors, including the design and operation of the multileaf collimator and associated backup diaphragms, the patterns of modulation specified by the treatment planning process and the algorithms used to convert the patterns of modulation into a series of leaf sequences by which the beams are delivered.
Starting with the region outside the boundary of the primary collimator, the leakage requirements set out in IEC 60601-2-1 [2] limit leakage to 0.1% of the dose on the central axis from an unmodulated beam. Thus, for a dose of 2 Gy on the central axis, the dose due to leakage will be less than 2 mGy. We can consider the effect of beam modulation on the actual dose resulting from leakage radiation. In cases where the beam is modulated by the insertion of either a physical wedge or a physical compensator, the actual dose due to leakage radiation will be increased by the wedge factor or compensator factor. If the wedge factor is 2.5, the dose due to leakage would increase from 2 mGy to 5 mGy. In the case of beam modulation by IMRT, using a programmed sequence of beam shapes or leaf trajectories an IMRT factor, analogous to the wedge factor or compensator factor, can be defined. This is the ratio of monitor units required to deliver the modulated field to the monitor units required to deliver the same dose to an unmodulated field of the same size and shape. The IMRT factor is always greater than unity and depends on the degree of modulation required and the particular sequences of leaf settings chosen to deliver the modulation. Webb [4] has shown that the problem of segmenting modulated beams into leaf sequences does not have a unique solution, and thus there are many ways of delivering the same modulation. This is obvious in the special case of a uniform beam that can be delivered by scanning a narrow window across the beam aperture, in which case a large number of monitor units would be required, or by scanning a wider window, in which case fewer monitor units would be required. It is clearly desirable to choose a sequence that requires the minimum number of monitor units for a given modulation. Such a sequence will be delivered in the shortest time and will result in the lowest dose due to leakage radiation. It is also important to recognize that the IMRT factor will in many cases be smaller than the wedge factor that would be required if IMRT was not available. This is because physical wedges are designed to cover a range of field sizes, with the wedge factor being defined by the largest field size to be used. Production of a wedge profile by programmed collimation is equivalent to designing a wedge for each field and, as long as efficiency is one of the design criteria, the IMRT "wedge factor" will be less than the normal wedge factor required for comparable modulation. It is also worth noting that dynamically created wedges and compensators will produce less scattered X-rays and less electron contamination than physical wedges and physical compensators. This contamination radiation is likely to be as big a problem as leakage radiation.
Just as a steep wedge has a high wedge factor, a highly modulated IMRT sequence will have a high IMRT factor. High wedge or IMRT factors and consequent dose from leakage radiation is, however, not the only reason for minimizing the modulation of beams. In uniform beams, dosimetric and geometric accuracy are essentially independent within the planning target volume (PTV). In the penumbra regions small errors in position of the beam edge obviously result in large errors in dose, but these regions are, by definition, outside the PTV. If highly modulated fields are used, then careful consideration has to be given to all aspects of geometric accuracy; it has been suggested that tolerances on the positioning of beam limiting devices should be reduced from 2 mm to 1 mm (or less). Recognition and avoidance of this set of problems rather than that of increased dose from leakage radiation will probably be the main reason for using the minimum modulation necessary to achieve the required dose distributions.
The design of the multileaf collimator (MLC) leaves and the use of backup or standard collimators will also be important. For MLC leaves, mechanical restrictions result in areas between the leaves where radiation can pass with less attenuation. The manufacturers have reduced this by introducing tongue and groove designs or by curving the edges of the leaves. This reduces the peak leakage to less than 5%, with an average leakage between 0.5% and 2.5%. This is still larger than that permitted for standard collimators. Different manufacturers have tried to address this in various ways, including the use of additional backup collimators [5], designing the MLC as an add-on device to a standard treatment head [6] or by relying solely on the MLC leaves and precise engineering to attenuate the beam and minimize leakage [7]. The former two methods reduce, outside the boundary of the backup collimators, the leakage to levels less than those of a standard collimator. Where there are no backups, then an area of higher leakage will extend over the whole width of the field. Thus, when leakage and IMRT are being considered, the role of the backup collimators is important. Where backup collimators are present, they should ideally always be placed at the position of the outer leaf, thus minimizing leakage. This can be achieved within the interpreter or sequencer module in which intensity maps are converted into a series of leaf positions. Where backups are not present, individual purchasers should consider this in the procurement process.
As well as technological changes, there are likely to be changes to clinical practice that will also impact, although on a smaller scale, on the integral dose to the patient outside the treatment field. The aim of conformal therapy and the use of IMRT is to reduce morbidity and, via dose escalation, to improve local control and ultimately survival. Dose escalation studies using IMRT are currently delivering doses of over 80 Gy [8]. Dose escalation may in future be addressed by changes in fractionation schedules, but the current practice of adding more equal sized fractions to the end of a standard treatment will invariably lead to increased integrated leakage radiation dose received by the patient. These increases will be much smaller than any increase due to changes in delivery technique. Individual patient dose escalation will not affect room shielding requirements since the same number of beam exposures will be delivered per day.
Conformal therapy and IMRT will also generally result in smaller volumes being treated. This will have an impact on the scattered radiation dose, which, as shown by Lillicrap et al [1], can be the dominant source of radiation outside the beam up to 14 cm from the beam edge in the example shown. The magnitude of scattered radiation is reduced for smaller conformal fields [9]. This decrease will act to off-set any increase in leakage radiation dose.
The effect of leakage radiation and increased beam on-time has also been raised regarding treatment room design [3]. Clearly, if the machine is going to be on for longer times per patient, then the duty cycle, i.e. the number of hours radiation is produced per 8 h day, will change. However, we believe that this will not change significantly owing to the logistics of radiotherapy delivery. Treatment of 40 patients in an 8 h day represents reasonably good productivity with the typical case mix for most radiotherapy departments. As treatments become more complex, needing, for example, increased numbers of beams, additional immobilization, in vivo dosimetry and portal imaging, patient through-put will be reduced. These factors are concerned with set up and verification rather than the delivery of the treatment itself. An integrated approach to the overall radiotherapy process will minimize the reduction in efficiency but not eliminate it. If IMRT is one of the factors contributing to increased complexity and if the delivery of the modulated fields is not optimized, there will be a further increase in the irradiation time, which will have an impact on the number of patients treated. Typical conventional treatments require the beam to be on for approximately 10% of the overall treatment time. All other things being equal, an IMRT (or wedge) factor of 9 would result in the beam being on for 45% of the overall time (Table 1
). It can be seen from this extreme example that the increase in beam on-time, which needs to be considered in the design of the machine and the room in relation to environmental protection, is less than predicted by simple application of the IMRT factor. At the present time there is a general consensus that particularly complex IMRT regimens require additional verification to be carried out. This increase in time per patient without the beam on will be as much, relatively, as the increase in beam on-time per patient. The net result will be to limit change in the relative beam-on to beam-off time for an 8 h day (fewer patients will be treated) and hence reduce the need for additional protection.
|
Received for publication October 23, 2000. Accepted for publication November 27, 2000.
References
This article has been cited by other articles:
 |
|
 |
|
  A. G. Rijkee, J. Zoetelief, C. P. J. Raaijmakers, S. C. Van Der Marck, and W. Van Der Zee Assessment of induction of secondary tumours due to various radiotherapy modalities Radiat Prot Dosimetry, May 1, 2006; 118(2): 219 - 226. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E J Hall and C-S Wuu BJR Review of the Year 2004 Br. J. Radiol., July 1, 2005; 78(931): 672 - 673. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| BJR | DMFR | IMAGING | ALL BIR JOURNALS |
