British Journal of Radiology 74 (2001),249-254 © 2001 The British Institute of Radiology
Influence of a vac-fix immobilization device on the accuracy of patient positioning during routine breast radiotherapy
C A Nalder1,
A M Bidmead1,
C D Mubata1,
D Tait2 and
C Beardmore,2
1 Departments of Physics
2 Radiotherapy and Oncology, Royal Marsden NHS Trust, Fulham Road, London SW3 6JJ, UK
 |
Abstract
|
|---|
Continued use of basic planning and treatment techniques, in contrast to the improved methods implemented at many other anatomical sites, has emphasized the need for improved breast dosimetry. Any future technique delivering a superior three-dimensional dose distribution will be of maximum benefit if set-up errors are minimized. To determine the influence of vacuum moulded bag (vac-fix) immobilization on routine breast radiotherapy, 17 patients received half their radiotherapy fractions using our standard breast board technique and half using a vac-fix device positioned on the breast board. Treatment accuracy and reproducibility were assessed for each technique using daily electronic portal imaging and were analysed in terms of random and systematic translational and rotational displacements of treatment fields with respect to corresponding simulation images. In addition, patients completed a short questionnaire aimed at determining which technique they preferred. Results showed that random errors for the two techniques did not differ significantly. Approximately 80% of random translations recorded were less than 3 mm and 80% of random rotations were less than 1.5°. Systematic errors showed some improvement with the vac-fix system. In the anteroposterior direction, approximately 80% of systematic errors were less than 4 mm for both techniques, but in the superoinferior direction the 80% point was reduced from 5.0 mm for the standard set-up to 2.7 mm for treatment in vac-fix. For rotational systematic errors, the corresponding value dropped from 1.8° for the standard set-up to 1.1° in vac-fix. Therefore, for many patients, additional use of a vac-fix device improved the transfer of the planned set-up from simulator to treatment unit. Additionally, answers to the questionnaire indicated that patients generally favoured the vac-fix system over use of the breast board alone. In conclusion, however, introduction of vac-fix immobilization for all patients was not thought justified as the improvements demonstrated are not likely to be clinically significant with the present treatment technique.
|
Introduction
|
|---|
Treatment of early breast cancer accounts for a substantial proportion of radiotherapy courses delivered in the UK. With simple planning techniques, the dose distributions achieved can be considerably non-uniform, with some regions of the breast receiving 10–30% more than the prescribed dose [1, 2]. This is especially true for women with larger breasts and it is likely that this is a major cause of the increased complications reported in this patient group [3–5].
Developments leading to improvements in three-dimensional dose distributions, such as the use of conformal radiotherapy and intensity modulated radiotherapy (IMRT), will be of maximum benefit if patient set-up errors can be minimized. Because of the tighter requirements for accuracy with these techniques, the importance of patient set-up errors has recently attracted considerable attention. Many studies have reported on the accuracy and reproducibility of patient set-up during treatment [6–10], while others have demonstrated improvements resulting from the use of immobilization devices [11, 12]. However, most of this work has dealt with pelvic radiotherapy. The aim of this study was to determine whether a significant improvement in set-up accuracy and reproducibility could be demonstrated during routine breast radiotherapy using vacuum moulded (vac-fix) bags for immobilization. Vac-fix bags were used in conjunction with a purpose-built breast board. The additional immobilization involved only a slight modification to the routine set-up with the aim that simulation and treatment times would not be significantly extended.
|
Method
|
|---|
Patients prescribed radiotherapy to the breast in 25 daily fractions over 5 weeks, following conservative surgery for early breast cancer, were eligible for entry into the trial. 17 patients were recruited and treated between January 1997 and December 1998. They received half their radiotherapy fractions using our standard breast set-up and half with additional vac-fix immobilization. Each patient therefore acted as her own control. To minimize the possible effects of time trends on the results [6], all courses of radiotherapy were divided into four sections or blocks of six fractions each, with the blocks alternating between the two techniques. The final fraction was given with whichever technique was used for the last treatment block. The initial technique and thus the sequence order of the two techniques was randomly assigned for individual patients.
Simulation and dosimetry were performed independently for both techniques, requiring all patients to attend two separate simulation sessions. Other than the presence or absence of the vac-fix bag, the simulation procedures were identical in all respects.
At this centre, radiotherapy to the breast only is performed using a typical symmetric two-field tangential technique. Gantry angles are chosen to achieve a non-divergent posterior beam edge, while couch and collimator rotations ensure a vertical superior plane for ease of matching future fields if necessary. The breast board (Figure 1
) was designed in-house and is well established in routine use. It allows adjustment of board angle and arm pole position and is used in conjunction with an adjustable foot rest. For simulation, the patient is positioned supine on the device and the board angle is chosen such that the sternum is parallel to the couch top. The arm on the treatment side is abducted to 90° with the patient holding a hand grip on the arm pole. The foot rest is positioned to restrict movement along the board. The head is kept straight. Once the treatment area is localized, simulator films of the tangential fields are taken and an outline through the central axis is measured. Anterior midline and lateral reference points are tattooed and the field borders marked on the patient. To avoid confusion on set, marks for vac-fix and non-vac-fix treatments were colour coded. Before commencement of the third and fourth treatment blocks, all patients made a brief return to the simulator so that the alternative set-up could be re-screened and the marks refreshed.
View larger version (109K):
[in this window]
[in a new window]
|
Figure 1. The breast board used for routine treatment at The Royal Marsden Hospital. Board angle and arm pole positions are adjustable.
|
|
The vac-fix system consists of a plastic, air-tight cushion filled with small expanded polystyrene spheres. The device is initially flexible, but on removal of air it provides a firm support, contoured to the shape of the individual patient. Rigidity should be maintained over the course of a typical treatment, but a bag can be periodically re-evacuated on set if necessary. In the case of simulation with vac-fix immobilization, the patient was laid back onto a partially evacuated bag positioned on top of the breast board. Again, the angle of the board was selected to position the sternum parallel to the couch. The bag was then fully evacuated while being shaped to provide support under the patient's back and shoulders. An example of an evacuated bag positioned on the breast board, as used in this study, is shown in Figure 2. Care was taken to ensure that the bag did not obscure the field border incident on the skin on the patient's treatment side. Marks were drawn on the bag corresponding to the patient's lateral reference points, enabling verification of patient position within the device during treatment.
Once on treatment, both tangential fields were imaged daily using an electronic portal imaging system. Images were acquired at the beginning of each fraction and, although not necessarily representative of the whole fraction, it was assumed that movement during the irradiation of each field is small [13] and less significant than set-up variation between fractions [14]. Each image was compared with its corresponding simulator reference, a digital version of which was stored in the portal imaging system. Before a software comparison can be made between the reference and portal images, certain features within the reference image are outlined to act as "match structures". To be useful, the match structures highlighted must be features that will be visible on the portal image. For tangential breast images, the only appropriate feature is the chest wall/lung interface. Once a match template is produced, the software will attempt to align the match structures with features in the portal image. If there is a discrepancy, the match template can be aligned with the portal image by eye. Once aligned, values for translational and rotational displacement of the treatment field with respect to the reference image are determined. It has been shown that this type of field matching process can be subject to relatively large systematic errors between different observers [15]. In the present study, all defining of match structures and image matching was carried out by a single observer to eliminate this source of error. Matching was performed for each block of vac-fix and non-vac-fix immobilization and results were compared to assess the performance of each system for individual patients and for the group of patients as a whole.
After each treatment block, patients were requested to fill in a simple questionnaire to determine which technique they favoured. Patients were not asked to compare the two techniques directly but rather to score aspects of each on a four-point scale. Two questions were asked: (a) how easy did you find it to maintain the treatment position over the last treatment block? And (b) how comfortable did you find the treatment position over the last treatment block? Responses were selected from the categories (1) not at all, (2) slightly, (3) moderately and (4) very.
Finally, a comparison of treatment parameters such as field length, field width, gantry angle and amount of lung in the field was made to see whether these were in any way affected by the presence of vac-fix immobilization.
|
Results
|
|---|
For each patient, translational and rotational deviations of the treatment field from the simulator reference were plotted against fraction number. Medial and lateral fields were treated separately. A general inspection of the data revealed no evidence of time trends; however, systematic shifts in field position between vac-fix and non-vac-fix treatment blocks were obvious in many patients. For some patients systematic shifts were also noted, although less frequently, between the two apparently identical treatment blocks. Owing to these differences, measurements from the different treatment blocks have been considered separately.
The total number of images analysed was 592 out of a possible 850. The missing images resulted from a combination of accelerator and imager routine service or breakdown. 304 images were of patients in the vac-fix device and 288 represented the standard technique. In all, 136 treatment blocks were analysed, 68 for each technique. The sign convention used derives from the image analysis software and is explained in Table 1. Table 2 shows the average systematic error for all blocks of a particular technique. The standard deviation gives a measure of the range of these systematic errors. Table 3 lists the standard deviation of the random errors for all patients for each technique.
View this table:
[in this window]
[in a new window]
|
Table 1. Sign convention from image analysis software. The direction indicated represents the shift of the treatment field with respect to the patient
|
|
In the anteroposterior direction there is no improvement demonstrated by additional use of a vac-fix support. Figures 3 and 4 are cumulative frequency distributions showing how the magnitude of the systematic and random errors are distributed for the two techniques.
In the superoinferior direction and for rotational deviations, use of the vac-fix bag was associated with a reduction in the treatment block systematic errors. Figures 5–8 illustrate cumulative frequency distributions for the systematic and random errors. Examination of the data revealed large improvements in individual patients, with little or no benefit and occasionally a deterioration for others. Table 4 compares point values from all the frequency distributions.
View this table:
[in this window]
[in a new window]
|
Table 4. Comparison of point values from the cumulative frequency distributions. The numbers indicate the magnitude of the errors (in mm or degrees) below which 50% and 80% of the recorded values lie
|
|
Comparison of the field length, field width, gantry angle and maximum lung depth in each field from simulation of the same patient using the two techniques revealed no systematic differences among the values.
For the patient questionnaires, the percentage of replies in each category for the two techniques is illustrated in Figures 9 and 10. In total, 47 questionnaires were available for analysis, 26 for vac-fix blocks and 21 for the standard technique. Overall, the response to both techniques was positive, with the majority of answers (90%) in the two most favourable categories. However, in both cases the vac-fix technique scored the highest percentage in the top category. Regarding maintenance of treatment position, 58% of responses considered this to be very easy using the vac-fix bag as opposed to 38% using the standard set-up. For patient comfort, 61% of responses rated the vac-fix device very comfortable vs only 24% for the standard technique.
|
Discussion and conclusions
|
|---|
This study evaluated the introduction of vac-fix bags into routine breast radiotherapy as a relatively minor modification of the standard set-up. It also served as a thorough assessment of the present technique. For all the parameters measured, random errors associated with the addition of the vac-fix bag showed no reduction compared with those for the standard technique. Therefore, as far as day-to-day set-up variability is concerned, no advantage was demonstrated. Additionally, although the questionnaire responses implied that the vac-fix system was generally preferred by patients, the standard treatment did not receive unfavourable criticism.
Other than in the anteroposterior direction, where no difference was encountered, systematic errors between simulation and treatment were generally reduced with the vac-fix system. It has been shown elsewhere [16] that the major component of error often lies in the transfer of a set-up from simulator to treatment unit and, in this respect, use of the vac-fix system was shown to be of benefit. For a small subset of patients, the improvements were particularly significant and would have implications for field matching in treating supraclavicular and axilliary nodes or when including an internal mammary field. The patients entering the trial covered a range of both patient size and breast size, but from the sample available no obvious criteria could be detected to allow prediction of the group likely to benefit. As the improvements are limited to the systematic component of the error, and as the random errors are comparable, errors could be eliminated by the adjustment of patient set-up using regular portal imaging. However, it is undoubtedly preferable to avoid introduction of a systematic error in the first place.
On the strength of these results and with the simple breast dosimetry presently applied at the Royal Marsden NHS Trust, introduction of vac-fix immobilization for all patients undergoing breast radiotherapy would not be justified, since the improvements demonstrated are not likely to be clinically significant. However, any such decision should be based on a rigorous assessment of the accuracy and reproducibility achieved with the local treatment technique. With regard to the introduction of more complex techniques, such as the use of missing tissue/lung compensation [17] or other forms of IMRT, it is envisaged that vac-fix immobilization will be used. Treatments of this type should benefit from improving the transfer of the planned patient position from simulation to treatment unit, especially if the information used to optimize dosimetry is collected at simulation rather than via portal imaging. In addition, improved comfort during treatment could assist patients in maintaining the treatment position over possibly extended treatment times. However, future requirements for CT planning may necessitate a complete redesign of the current breast technique.
|
Footnotes
|
|---|
Current address for C Beardmore: Department of Radiotherapy, Royal Berkshire Cancer Centre, London Road, Reading RG1 5AN, UK. 
Received for publication February 9, 2000.
Revision received September 21, 2000.
Accepted for publication October 20, 2000.
|
References
|
|---|
-
Chu JCH, Sontag MR, Danoff BF, Goodman RL. Assessment of dose uniformity in the irradiation of the intact breast using three-dimensional dose calculations. Int J Radiat Oncol Biol Phys 1985;11(Suppl.):163
-
Solin LJ, Chu JCH, Sontag MR, Brewster L, Cheng E, Doppke K, et al. Three-dimensional photon treatment planning of the intact breast. Int J Radiat Oncol Biol Phys 1991;21:193–203.[Medline]
-
Neal AJ, Torr M, Helyer S, Yarnold JR. Correlation of breast dose heterogeneity with breast size using three dimensional CT planning and dose volume histograms. Radiother Oncol 1995;34:210–8.[Medline]
-
Moody AM, Mayles WPM, Bliss JM, A'Hern RP, Regan J, Broad B, et al. The influence of breast size on late radiation effects and association with radiotherapy dose inhomogeneity. Radiother Oncol 1994;33:106–12.[Medline]
-
Neal AJ, Mayles WPM, Yarnold JR. Tangential breast irradiation—rationale and methods for improving dosimetry. Br J Radiol 1994;67:1149–54.[Abstract]
-
El-Gayed AAH, Bel A, Vijlbrief R, Bartelink H, Lesbesque JV. Time trend of patient set-up deviations during pelvic irradiation using electronic portal imaging. Radiother Oncol 1993;26:162–71.[Medline]
-
Griffiths SE, Khoury GG, Eddy A. Quality control of radiotherapy during pelvic irradiation. Radiother Oncol 1991;20:203–6.[Medline]
-
Van Tienhoven G, Lanson JH, Crabeels D, Heukelom S, Mijnheer BJ. Accuracy in tangential breast treatment set-up: a portal imaging study. Radiother Oncol 1991;22:317–22.[Medline]
-
Pradier O, Schmidberger H, Weiss E, Bouscayrol H, Daban A, Hess CF. Accuracy of alignment in breast irradiation: a retrospective analysis of clinical practice. Br J Radiol 1999;72:685–90.[Abstract]
-
Creutzberg CL, Althof VG, Huizenga H, Visser AG, Levendag PC. Quality assurance using portal imaging: the accuracy of patient positioning in irradiation of breast cancer. Int J Radiat Oncol Biol Phys 1993;25:529–39.[Medline]
-
Soffen EM, Hanks GE, Hwang CC, Chu JCH. Conformal static field therapy for low volume low grade prostate cancer with rigid immobilisation. Int J Radiat Oncol Biol Phys 1991;20:141–6.[Medline]
-
Rosenthal SA, Roach M, Goldsmith BJ, Doggett EC, Pickett B, You HS, et al. Immobilisation improves the reproducibility of patient positioning during six-field conformal radiation therapy for prostate carcinoma. Int J Radiat Oncol Biol Phys 1993;27:921–6.[Medline]
-
Fein DA, McGee KP, Schultheiss TE, Fowble BL, Hanks GE. Intra- and interfractional reproducibility of tangential breast fields: a prospective on-line portal imaging study. Int J Radiat Oncol Biol Phys 1996;34:733–40.[Medline]
-
Westbrook C, Gildersleve J, Yarnold JR. Quality assurance in daily treatment procedure: patient movement during tangential field treatment. Radiother Oncol 1991;22:299–303.[Medline]
-
Bijhold J, van Herk M, Vijlbrief R, Lebesque JV. Fast evaluation of patient set-up during radiotherapy by aligning features in portal and simulator images. Phys Med Biol 1901;36:1665–79.
-
Das IJ, Cheng CW, Fosmire H, Kase KR, Fitzgerald TJ. Tolerances in setup and dosimetric errors in the radiation treatment of breast cancer. Int J Radiat Oncol Biol Phys 1993;26:883–90.[Medline]
-
Mayles WPM, Yarnold JR, Webb S. Improved dose homogeneity in the breast using tissue compensators. Radiother Oncol 1991;22:248–51.[Medline]
This article has been cited by other articles:
|
|
|
|
 
S Chopra, K A Dinshaw, R Kamble, and R Sarin
Breast movement during normal and deep breathing, respiratory training and set up errors: implications for external beam partial breast irradiation.
Br. J. Radiol.,
September 1, 2006;
79(945):
766 - 773.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
P T Truong, E Berthelet, V Patenaude, J Bishop, B Sandwith, V Moravan, W Beckham, T Mitchell, and I A Olivotto
Setup variations in locoregional radiotherapy for breast cancer: an electronic portal imaging study
Br. J. Radiol.,
August 1, 2005;
78(932):
742 - 745.
[Abstract]
[Full Text]
[PDF]
|
|
|
Top
Abstract
Introduction
) and with vac-fix immobilization (
).
, with vac-fix immobilization; 
, no vac-fix immobilization.