- Email: [email protected]
Is one head and neck immobilization system as good as another? One center’s experience
Medical Dosimetry, Vol. 28, No. 1, pp. 39 – 43, 2003 Copyright © 2003 American Association of Medical Dosimetrists Printed in the USA. All rights reserved 0958-3947/03/$–see front matter
doi:10.1016/S0958-3947(02)00240-6
IS ONE HEAD AND NECK IMMOBILIZATION SYSTEM AS GOOD AS ANOTHER? ONE CENTER’S EXPERIENCE LEAH LORD, BAPP, SCI (M.R.T.), SHARON MAY, DIP.F.M.I., MICHAEL BAILEY, M.SC. (STATISTICS)*, and LEIGH SMITH, B.H.A. William Buckland Radiotherapy Centre, The Alfred, Melbourne, Australia; and *Monash University, Department of Epidemiology and Preventative Medicine, Melbourne, Australia (Accepted 1 June 2002)
Abstract—The William Buckland Radiotherapy Center has used 2 different immobilization systems for patients requiring radiotherapy to the head-and-neck region. A polycarbonate mask was manufactured for radical treatments and a thermoplastic mask for palliative treatments. This study evaluated field placement accuracy, staff opinion, and production costs of both systems. The manual matching program of Varian PortalVision Electronic Portal Imaging (EPI) System was used to assess field placement accuracy on a daily basis. Radiation therapists (RTs) were surveyed before and after the study to determine their opinions of each system. Production time and required materials were recorded to assess cost. Nineteen patients from each system had daily EPI results compiled with no statistically significant difference observed in field placement accuracy. The thermoplastic system was found to be more cost efficient due to a combination of the reduced production time and reuseability of the masks. User acceptability of the thermoplastic system has increased so that it is now the preferred system. In conclusion, the thermoplastic system is a viable alternative to the polycarbonate system in terms of treatment accuracy and cost. It is recommended that the thermoplastic system be used for all radical and palliative treatments. In addition, RTs prefer the thermoplastic system. © 2003 American Association of Medical Dosimetrists. Key Words:
Radiotherapy, Immobilization, Head and neck, Electronic portal imaging, Dosimetrists.
material is thermoplastic, using brands such as Efficast (Orfit Industries, Belgium), Aquaplast (Aquaplast Corporation, New Jersey), and Orfit (Orfit). The mask is heated in a water bath and applied directly to the patient. This latter process eliminates 1 patient visit to the department, as an impression needed for a plaster positive is not required. A literature search found only 1 other study, by Weltons et al., that assessed 2 different systems within 1 department for both accuracy and cost.5 However, Weltons et al.5 used mainly port films (85%) to assess field placement and did not investigate radiation therapist (RT) opinion. Other published studies focused only on polycarbonate masks for field placement accuracy and used weekly electronic portal images (EPI) and bony anatomy6 or port films7 for the assessment. Thermoplastic immobilization systems have been compared with a bite-block system8 or a strip system2 for field placement accuracy alone. The William Buckland Radiotherapy Centre (WBRC) initially used a polycarbonate (Vivak) material for all treatments when the center opened in 1992. An alternative thermoplastic (Efficast) material has been available for use from mid-1997 and has been extensively trialed at WBRC for patients undergoing palliative treatments. The aim of this study was to determine whether the thermoplastic system is a viable alternative to the poly-
INTRODUCTION During the treatment of head-and-neck tumors, patients’ involuntary head movements during irradiation have proven to be a significant source of error.1 It is necessary to immobilize these patients for treatment so that radiation beams can be directed accurately to the same target area on a daily basis. Careful immobilization is essential in the head-and-neck area, where the tumor is often located adjacent to radiosensitive structures and the margins surrounding the tumor are small.2,3 Historically, immobilization techniques have varied widely, from a simple piece of tape placed over the forehead,3 or a head cup and neck roll,4 to a customized face mask.3 There are 2 different types of customized mask materials currently used in radiotherapy departments. The first type is a polycarbonate material using plastics such as Kablite (Mediplex, Epping, NSW, Australia), and Vivak (Menzel Plastic Traders, Dingley, Victoria, Australia) as a raw material. The polycarbonate plastic is made into an immobilization mask by using a vacuum former to heat the plastic and then form it around a plaster positive of the patient’s face. The second mask Reprint requests to: Leah Lord, William Buckland Radiotherapy Centre, The Alfred Hospital, Commercial Road Prahran, 3181 Victoria, Australia. E-mail: [email protected] Presented in part at the 1999 Australian Institute of Radiography Conference, March 17–19, 1999, Melbourne, Victoria, Australia; and the 1999 NZMRT Conference, 26 –28 August, 1999, Rotorua, New Zealand. 39
40
Medical Dosimetry
carbonate system in terms of treatment accuracy, cost, and user acceptability. MATERIALS AND METHODS Patient characteristics Patient eligibility was determined based on a minimum number of 10 fractions, from which there were at least 10 consecutive EPIs available for matching. Current work practices were followed where radical patients had polycarbonate masks manufactured and palliative patients, thermoplastic masks. All patients had routine field images captured during treatment exposure. The field placement assessment was completed retrospectively from archived EPIs. Field placement accuracy Similar studies have determined field placement accuracy in a number of ways, including the assessment of weekly EPIs9 or port films. Many studies have used port films with a frequency that varies from daily,10 to weekly,2 to 1 taken at the start of treatment and then 4 to 5 weeks into treatment,11 or the first and final session of the treatment course.12 One such study involved 9 patients returning to the simulator on a weekly basis and a comparative film taken.13 Methods of quantifying accuracy include using bony anatomy to measure movement,1,6,7 comparing distances between anatomical structures and field edges utilizing a scale,12 gluing 1 mm diameter lead shots to the mask for reference points,14 or using landmarks on the skin.8 Despite the immobilization used in most of these studies, deviations of up to 3 mm were found.10 EPIs’ are a method of treatment verification that allow the immediate acquisition of treatment images. Routine daily EPI’s obtained in this study for both immobilization systems were collected for analysis of field placement accuracy against the simulator film. Anatomical structures, bone, and soft tissue on the simulator film were outlined to form a template using the matching tools available in the Varian PortalVision program. The program then overlays the template and matches field edges on to the daily-acquired EPI. The user adjusts the template to match the anatomy on the simulator film. Any deviations in the lateral, longitudinal, and rotation directions from the simulator film are displayed in millimeters and are documented. The rotation was recorded but not assessed when it was noted that variation in rotation remained within ⫾ 1°. Palliatively treated patients tended to have fewer fractions than radical treatments. Therefore, only the first 10 fractions of each course of treatment from each mask were assessed and compared. One RT matched all of the images, while a second RT independently matched 31% of the total images to ensure reliability of the results. A correlation coefficient was applied to assess reliability. Results of the daily
Volume 28, Number 1, 2003
matching required agreement to within ⫾ 3 mm, otherwise the process was reviewed and the anatomy used was compared to determine if agreement could be reached. Statistical analysis Data was analyzed using SAS Version 6.12.15 As both lateral and longitudinal deviations were found to be normally distributed, data was analyzed using repeated measures analysis or variance. A p-value of 0.05 was considered to be statistically significant. Costs Cost can be an important factor when assessing whether to introduce any new system. Unit cost calculations included raw material, accessories, and associated labor. For the purpose of this study, the fixed costs, such as baseboards and neck shapes, were ignored, as they represent a similar cost for both systems. The raw materials of each system are available in 2 different sizes, small and large. The small size is used to treat the brain and the larger size is used to treat the head-and-neck area. The labor cost was calculated by multiplying the average time to produce a mask by the hourly wage of a grade 1, year 6 RT. The average time includes all patient contact for the impression and fitting plus the RT time in manufacturing and producing the mask. Fifteen mask production times were recorded for each system. The unit production cost (UPC) was calculated by adding the labor cost to the raw material cost. The thermoplastic unit production cost equation was calculated by adding the labor cost to a quarter of the raw material, because thermoplastic masks can be used at least 4 times before elastic properties are lost. User acceptability of the thermoplastic vs. polycarbonate systems Two surveys were completed to ascertain staff opinion of the thermoplastic system. The first survey was undertaken shortly after the introduction of the thermoplastic system into routine clinical use. The second survey was undertaken 2 years later. The RT staff in the planning and treatment areas were surveyed. The systems were compared for a number of different parameters and utilized a sliding scale, allowing choices ranging between poor and excellent for both types of mask. The parameters assessed were: ease of construction, production time, achieving the required position, marking the mask, using the mask for CT planning and in simulator, fitting the mask, clinical viewing of anatomy, ease of adjustment, mask stability, field adjustment, and field placement accuracy. RESULTS Field placement accuracy A total of 38 patient comparisons was completed, which included 19 polycarbonate and 19 thermoplastic
Comparison of two immobilization systems ● L. LORD et al.
Fig. 1. Lateral deviation from the isocenter for thermoplastic and polycarbonate treatments.
masks. EPI daily image assessments were performed on a total of 652 images. One RT matched all of the images using the EPI tools and an independent observer assessed 31% of the matched images. The lateral correlation coefficient achieved between the 2 RTs was 0.70 and the longitudinal correlation coefficient was 0.76. The coefficient has a positive correlation and validates the reliability of the results between the 2 RTs. The lateral deviation in the matching program includes both right-left and anterior-posterior movement, while the longitudinal deviation indicates superior-inferior movement. No clinically significant differences were found between the groups studied for the rotation results. The results, shown in Figs. 1 and 2, indicate that both the longitudinal and lateral deviations follow a normal distribution, which is similar to the distribution seen by Hunt et al.10 As can be seen in Table 1, the field placement accuracy range of deviations for the polycarbonate system was larger than that of the thermoplastic system. The means were very similar, as were the median and standard deviation. The least-squares mean assessed the absolute mean deviation from the planned isocenter. As can be seen, the thermoplastic system mean deviation was slightly less than the polycarbonate system. Statistical analysis of the data has shown no significant difference in accuracy between the polycarbonate and thermoplastic systems.
Fig. 2. Longitudinal deviation from the isocenter for thermoplastic and polycarbonate treatments.
41
Fig. 3. A comparison of the time taken to produce a thermoplastic mask to a polycarbonate mask.
Production costs Raw material costs. The raw material costs of both systems were obtained from the manufacturers. The polycarbonate material, together with the plaster and plaster bandages were cheaper to purchase than the thermoplastic material. For example, the raw material cost for a small polycarbonate sheet was one sixth the cost of a thermoplastic sheet of similar size and the cost of a large polycarbonate sheet was one seventh the cost of its equivalent thermoplastic sheet. Labor cost. The labor cost of the polycarbonate material was greater due to the labor-intensive process of forming the mask. Figure 3 is a comparison of the time taken to produce 15 polycarbonate and 15 thermoplastic masks. The average labor time to produce a polycarbonate mask was 126 minutes, while the average labor time to produce a thermoplastic mask was 13.6 minutes. The time taken to produce a thermoplastic mask at the WBRC was faster than Gerber et al.11 who quoted 20 to 30 minutes. Unit production cost. The UPC for the polycarbonate mask was $AUD51.32 for the small size and $AUD57.17 for the large size. The UPC for a thermoplastic mask was $AUD48.88 for the small size if used once or $AUD12.22 when used 4 times. A large mask was $AUD74.88 if used once or $AUD18.72 when used 4 times. The above results indicated that the thermoplastic small mask is approximately one quarter the cost of the equivalent polycarbonate mask, while a large thermoplastic mask is approximately one third the cost of its equivalent polycarbonate mask when used 4 times. The thermoplastic system, when re-used, represents a dramatic cost saving to a department. User acceptability of the thermoplastic vs. polycarbonate systems Surveys were distributed between July and August 1997 to 22 members of the planning and treatment RT staff. Thirteen surveys were returned (59%). Surveys were then redistributed between July and August 1999 to
42
Medical Dosimetry
Volume 28, Number 1, 2003
Table 1. Results of field placement accuracy using the Varian electronic portal imaging system Deviation Number of fields matched Range of deviations (mm) Mean (mm) Median (mm) Standard Deviation (mm) Least-squares mean Least-squares Standard Error
Thermoplastic Lateral
Polycarbonate Lateral
Thermoplastic Longitudinal
Polycarbonate Longitudinal
317.0 ⫺7.3–8.1 ⫺0.2 ⫺0.2 1.7 1.51 0.16
335.0 ⫺5.8–9.3 ⫺0.3 ⫺0.1 2.1 1.56 0.18
317.0 ⫺6.6–7.1 ⫺0.3 ⫺0.1 1.8 1.48 0.14
335.0 ⫺7.4–10 0.1 0.2 1.9 1.52 0.16
22 of the then-current planning and treatment RT staff. Seventeen surveys were returned (77%). The results of both surveys are shown in Table 2. The results of the initial survey indicated that RTs at the WBRC preferred the polycarbonate system. The main disadvantage of the thermoplastic system was its opacity. Anatomical landmarks such as tragal notch are routinely matched on the mask to the skin for mask fit. This procedure was difficult using the thermoplastic mask, and RTs perceived the lack of visual verification a distinct disadvantage. The repeat survey indicated a change in attitude toward the thermoplastic system. The main perceived disadvantages of the opacity of the thermoplastic mask was still a slight disadvantage. However, 1 advantage of the thermoplastic system, recognized as important by the RTs, was the benefit to the patient in requiring only 1 visit for mask production immediately presimulation. Also, the ease of construction of a thermoplastic mask reduced the production time to approximately one tenth that of a polycarbonate mask. DISCUSSION AND CONCLUSIONS When assessing field placement, radically-treated patients were compared with those receiving palliative treatments. While it may be thought the 2 groups are insignificantly different, it should be noted that the latter have underlying conditions that tend to make them more difficult to treat. These include swelling, pain, discomfort, and mobility problems. These factors often make them more difficult to stabilize. Despite the bias in the
palliative group of patients, the thermoplastic system achieved a comparable accuracy to the polycarbonate system. The results of this study are difficult to compare with other studies because of different methods of analysis. The results for the polycarbonate system masks can be compared with another field placement accuracy study that used EPIs.6 The mean deviation achieved by Bel et al.6 was ⫺0.8 mm laterally and ⫺0.8 mm longitudinally compared with the results of ⫺0.2 mm to ⫺0.1 mm achieved in this study. The standard deviation achieved by Bel et al.,6 although clinically comparable, was more accurate with a value of 1.5 mm laterally and 1.6 mm longitudinally compared to the WBRC results of 2.1 mm laterally and 1.9 mm longitudinally for polycarbonate masks. The intertreatment standard deviation achieved by Fiorino et al. for thermoplastic masks was less than 2 mm,16 which is comparable with the WBRC results. The absolute deviation from the isocenter was 2.38 mm for the thermoplastic system and 2.46 mm for the polycarbonate system and is comparable to the range cited by other studies1,11,12,14 and more complicated devices.13,17 When assessing which thermoplastic system to trial as an alternative to the polycarbonate system, a few alternate thermoplastic materials were chosen to determine which 1 was more suited to the department. The system that was chosen for its combination of malleability, reusability, and cost was the Biomet Efficast system. Fixed costs associated with each system were not con-
Table 2. Results of the initial and repeat surveys on RT staff opinion Category Ease of construction Time taken to construct Achieving required position Marking of cast In simulator Accuracy Viewing of anatomy Fit of cast Ease of adjustment Stability of cast Field adjustment
Initial Preference
Follow-up Preference
Opinion changed
Thermoplastic Thermoplastic Polycarbonate
Thermoplastic Thermoplastic Either
No No Yes
Polycarbonate Polycarbonate Polycarbonate Polycarbonate Either Either Either Either
Polycarbonate Polycarbonate Either Polycarbonate Thermoplastic Thermoplastic Either Either
No No Yes No Yes Yes No No
Comparison of two immobilization systems ● L. LORD et al.
sidered, as the analysis dealt with UPCs and were similar. Fixed cost items included baseboards, standard neck shapes, and vacuum bags. The main cost advantage of the thermoplastic system is its ability to be used at least 4 times. The WBRC is currently using the thermoplastic system for all patients requiring an immobilization shell. There are 2 groups of patients—larynx and electron treatment areas—that routinely have areas cut out of the mask. If thermoplastic masks have areas cut out, they lose their elastic properties and are unable to be reused. A mask on its third or fourth application is more cost effective to use for these types of treatment. Once used, the masks are washed in warm soapy water and dried flat, ready to be used for the next patient. If the mask has bodily fluids from the previous patient, it is discarded. When the thermoplastic system was initially introduced to the WBRC, user acceptance was low. There has been a gradual acceptance from the RTs by initially restricting use of the thermoplastic masks for palliative treatments. The main disadvantage of this system to our RTs is the masks’ opacity. After 2 years of routine clinical use and hands-on experience, the improved acceptance can be summed up by 1 survey comment stating “my attitude to the thermoplastic system has improved dramatically with increased use.” The recommendation to obtain full acceptance of the thermoplastic system would be to develop a translucent mask material. This study indicates that the initial perception of the RTs that the polycarbonate system is more accurate has changed. This change in attitude has been aided by the knowledge that the accuracy of the systems is comparable. There is no statistically significant difference between the 2 systems in terms of treatment accuracy. This study has shown the thermoplastic system to be more cost effective even when it is reused at least twice, although it can be used up to 4 times. The thermoplastic system has the additional advantage for the patient that it requires 1 less visit to the department and reduces the time dedicated to forming a mask by the staff. The initial recommendation was to use this thermoplastic system for all patients, excluding those patients whose masks are routinely cut out unless the mask material is on its final use. Four years after the thermoplastic masks have been introduced, they are now being used exclusively for all brain and head-and-neck patients.
43
Acknowledgments—The authors thank Kay Hatherly and Professor Alan Rodger for their editing skills. The authors also thank BioMet for their financial support in assisting to present this paper at the NZMRT conference.
REFERENCES 1. Niewald, M.; Lehmann, W.; Uhlmann, U.; et al. Plastic material used to optimise radiotherapy of the head and neck tumours and the mammary carcinoma. Radiother. Oncol 11:55–63; 1988. 2. Bentel, G.; Marks, L.; Hendren, K.; et al. Comparison of two head and neck immobilisation systems. Int. J. Radiat. Oncol. Biol. Phys 38:867–73; 1997. 3. Bentel, G.; Marks, L.; Sherouse, G.; et al. A customised head and neck support system. Int. J. Radiat. Oncol. Biol. Phys 32:245–8; 1995. 4. Rabinowitz, I.; Broomberg, J.; Goitein, M.; et al. Accuracy of radiation field alignment in clinical practice. Int. J. Radiat. Oncol. Biol. Phys 11:1857–67; 1985. 5. Weltens, C.; Kesteloot, K.; Vandevelde, G.; et al. Comparison of plastic and Orfit® masks for patient head fixation during radiotherapy: Precision and costs. Int. J. Radiat. Oncol. Biol. Phys 33:499 –507; 1995. 6. Bel, A.; Keus, R.; Vijlbrief, R.; et al. Set up deviations in wedged pair irradiation of parotid gland and tonsillar tumours measured with an electronic portal imaging device. Radiother. Oncol 37: 153–9; 1995. 7. Marsh, R.; Balter, J.; Evans, V.; et al. Design and analysis of an immobilisation and repositioning system for treatment of neck malignancies. Med. Dosim 22:293–7; 1997. 8. Sweeney, R.; Bale, R.; Vogele, M.; et al. Repositioning accuracy: Comparison of a non-invasive head holder with thermoplastic mask for fractionated radiotherapy and a case report. Int. J. Radiat. Oncol. Biol. Phys 41:475–83; 1998. 9. Hess, C.; Kortmann, R.; Jany, R.; et al. Accuracy of field alignment in radiotherapy of head and neck cancer utilising individualised face mask immobilisation: A retrospective analysis of clinical practise. Radiother. Oncol 34:69 –72; 1995. 10. Hunt, M.; Kutcher, G.; Burman, C.; et al. The effect of setup uncertainties on the treatment of nasopharynx cancer. Int. J. Radiat. Oncol. Biol. Phys 27:437–47; 1993. 11. Gerber, R.; Marks, J.; Purdy, J. The use of thermal plastics for immobilisation of treatments during radiotherapy. Int. J. Radiat. Oncol. Biol. Phys 8:1461–2; 1982. 12. Huizenga, H.; Levendag, P.; De Porre, M.; et al. Accuracy in radiation field alignment in head and neck cancer: A prospective study. Radiother. Oncol 11:181–7; 1988. 13. Thornton, A.; Ten Haken, R.; Gerhardsson, A.; et al. Threedimensional motion analysis of an improved head immobilisation system for simulation, CT, MRI and PET imaging. Radiother. Oncol 20:224 –8; 1991. 14. Dunscombe, P.; Fox, K.; Loose, S.; et al. The investigation and rectification of field placement errors in the delivery of complex head and neck fields. Int. J. Radiat. Oncol. Biol. Phys 26:155–61; 1993. 15. SAS Institute Inc., Carey, NC. 16. Fiorino, C.; Cattaneo, G.M.; Del Vecchio, A.; et al. Skin-sparing reduction effects of thermoplastics used for immobilisation in head and neck radiotherapy. Radiother. Oncol 30:267–70; 1994. 17. Verhey, L.; Goitein, M.; McNulty, P.; et al. Precise positioning of treatments for radiation therapy. Int. J. Radiat. Oncol. Biol. Phys 8:289 –94; 1982.
doi:10.1016/S0958-3947(02)00240-6
IS ONE HEAD AND NECK IMMOBILIZATION SYSTEM AS GOOD AS ANOTHER? ONE CENTER’S EXPERIENCE LEAH LORD, BAPP, SCI (M.R.T.), SHARON MAY, DIP.F.M.I., MICHAEL BAILEY, M.SC. (STATISTICS)*, and LEIGH SMITH, B.H.A. William Buckland Radiotherapy Centre, The Alfred, Melbourne, Australia; and *Monash University, Department of Epidemiology and Preventative Medicine, Melbourne, Australia (Accepted 1 June 2002)
Abstract—The William Buckland Radiotherapy Center has used 2 different immobilization systems for patients requiring radiotherapy to the head-and-neck region. A polycarbonate mask was manufactured for radical treatments and a thermoplastic mask for palliative treatments. This study evaluated field placement accuracy, staff opinion, and production costs of both systems. The manual matching program of Varian PortalVision Electronic Portal Imaging (EPI) System was used to assess field placement accuracy on a daily basis. Radiation therapists (RTs) were surveyed before and after the study to determine their opinions of each system. Production time and required materials were recorded to assess cost. Nineteen patients from each system had daily EPI results compiled with no statistically significant difference observed in field placement accuracy. The thermoplastic system was found to be more cost efficient due to a combination of the reduced production time and reuseability of the masks. User acceptability of the thermoplastic system has increased so that it is now the preferred system. In conclusion, the thermoplastic system is a viable alternative to the polycarbonate system in terms of treatment accuracy and cost. It is recommended that the thermoplastic system be used for all radical and palliative treatments. In addition, RTs prefer the thermoplastic system. © 2003 American Association of Medical Dosimetrists. Key Words:
Radiotherapy, Immobilization, Head and neck, Electronic portal imaging, Dosimetrists.
material is thermoplastic, using brands such as Efficast (Orfit Industries, Belgium), Aquaplast (Aquaplast Corporation, New Jersey), and Orfit (Orfit). The mask is heated in a water bath and applied directly to the patient. This latter process eliminates 1 patient visit to the department, as an impression needed for a plaster positive is not required. A literature search found only 1 other study, by Weltons et al., that assessed 2 different systems within 1 department for both accuracy and cost.5 However, Weltons et al.5 used mainly port films (85%) to assess field placement and did not investigate radiation therapist (RT) opinion. Other published studies focused only on polycarbonate masks for field placement accuracy and used weekly electronic portal images (EPI) and bony anatomy6 or port films7 for the assessment. Thermoplastic immobilization systems have been compared with a bite-block system8 or a strip system2 for field placement accuracy alone. The William Buckland Radiotherapy Centre (WBRC) initially used a polycarbonate (Vivak) material for all treatments when the center opened in 1992. An alternative thermoplastic (Efficast) material has been available for use from mid-1997 and has been extensively trialed at WBRC for patients undergoing palliative treatments. The aim of this study was to determine whether the thermoplastic system is a viable alternative to the poly-
INTRODUCTION During the treatment of head-and-neck tumors, patients’ involuntary head movements during irradiation have proven to be a significant source of error.1 It is necessary to immobilize these patients for treatment so that radiation beams can be directed accurately to the same target area on a daily basis. Careful immobilization is essential in the head-and-neck area, where the tumor is often located adjacent to radiosensitive structures and the margins surrounding the tumor are small.2,3 Historically, immobilization techniques have varied widely, from a simple piece of tape placed over the forehead,3 or a head cup and neck roll,4 to a customized face mask.3 There are 2 different types of customized mask materials currently used in radiotherapy departments. The first type is a polycarbonate material using plastics such as Kablite (Mediplex, Epping, NSW, Australia), and Vivak (Menzel Plastic Traders, Dingley, Victoria, Australia) as a raw material. The polycarbonate plastic is made into an immobilization mask by using a vacuum former to heat the plastic and then form it around a plaster positive of the patient’s face. The second mask Reprint requests to: Leah Lord, William Buckland Radiotherapy Centre, The Alfred Hospital, Commercial Road Prahran, 3181 Victoria, Australia. E-mail: [email protected] Presented in part at the 1999 Australian Institute of Radiography Conference, March 17–19, 1999, Melbourne, Victoria, Australia; and the 1999 NZMRT Conference, 26 –28 August, 1999, Rotorua, New Zealand. 39
40
Medical Dosimetry
carbonate system in terms of treatment accuracy, cost, and user acceptability. MATERIALS AND METHODS Patient characteristics Patient eligibility was determined based on a minimum number of 10 fractions, from which there were at least 10 consecutive EPIs available for matching. Current work practices were followed where radical patients had polycarbonate masks manufactured and palliative patients, thermoplastic masks. All patients had routine field images captured during treatment exposure. The field placement assessment was completed retrospectively from archived EPIs. Field placement accuracy Similar studies have determined field placement accuracy in a number of ways, including the assessment of weekly EPIs9 or port films. Many studies have used port films with a frequency that varies from daily,10 to weekly,2 to 1 taken at the start of treatment and then 4 to 5 weeks into treatment,11 or the first and final session of the treatment course.12 One such study involved 9 patients returning to the simulator on a weekly basis and a comparative film taken.13 Methods of quantifying accuracy include using bony anatomy to measure movement,1,6,7 comparing distances between anatomical structures and field edges utilizing a scale,12 gluing 1 mm diameter lead shots to the mask for reference points,14 or using landmarks on the skin.8 Despite the immobilization used in most of these studies, deviations of up to 3 mm were found.10 EPIs’ are a method of treatment verification that allow the immediate acquisition of treatment images. Routine daily EPI’s obtained in this study for both immobilization systems were collected for analysis of field placement accuracy against the simulator film. Anatomical structures, bone, and soft tissue on the simulator film were outlined to form a template using the matching tools available in the Varian PortalVision program. The program then overlays the template and matches field edges on to the daily-acquired EPI. The user adjusts the template to match the anatomy on the simulator film. Any deviations in the lateral, longitudinal, and rotation directions from the simulator film are displayed in millimeters and are documented. The rotation was recorded but not assessed when it was noted that variation in rotation remained within ⫾ 1°. Palliatively treated patients tended to have fewer fractions than radical treatments. Therefore, only the first 10 fractions of each course of treatment from each mask were assessed and compared. One RT matched all of the images, while a second RT independently matched 31% of the total images to ensure reliability of the results. A correlation coefficient was applied to assess reliability. Results of the daily
Volume 28, Number 1, 2003
matching required agreement to within ⫾ 3 mm, otherwise the process was reviewed and the anatomy used was compared to determine if agreement could be reached. Statistical analysis Data was analyzed using SAS Version 6.12.15 As both lateral and longitudinal deviations were found to be normally distributed, data was analyzed using repeated measures analysis or variance. A p-value of 0.05 was considered to be statistically significant. Costs Cost can be an important factor when assessing whether to introduce any new system. Unit cost calculations included raw material, accessories, and associated labor. For the purpose of this study, the fixed costs, such as baseboards and neck shapes, were ignored, as they represent a similar cost for both systems. The raw materials of each system are available in 2 different sizes, small and large. The small size is used to treat the brain and the larger size is used to treat the head-and-neck area. The labor cost was calculated by multiplying the average time to produce a mask by the hourly wage of a grade 1, year 6 RT. The average time includes all patient contact for the impression and fitting plus the RT time in manufacturing and producing the mask. Fifteen mask production times were recorded for each system. The unit production cost (UPC) was calculated by adding the labor cost to the raw material cost. The thermoplastic unit production cost equation was calculated by adding the labor cost to a quarter of the raw material, because thermoplastic masks can be used at least 4 times before elastic properties are lost. User acceptability of the thermoplastic vs. polycarbonate systems Two surveys were completed to ascertain staff opinion of the thermoplastic system. The first survey was undertaken shortly after the introduction of the thermoplastic system into routine clinical use. The second survey was undertaken 2 years later. The RT staff in the planning and treatment areas were surveyed. The systems were compared for a number of different parameters and utilized a sliding scale, allowing choices ranging between poor and excellent for both types of mask. The parameters assessed were: ease of construction, production time, achieving the required position, marking the mask, using the mask for CT planning and in simulator, fitting the mask, clinical viewing of anatomy, ease of adjustment, mask stability, field adjustment, and field placement accuracy. RESULTS Field placement accuracy A total of 38 patient comparisons was completed, which included 19 polycarbonate and 19 thermoplastic
Comparison of two immobilization systems ● L. LORD et al.
Fig. 1. Lateral deviation from the isocenter for thermoplastic and polycarbonate treatments.
masks. EPI daily image assessments were performed on a total of 652 images. One RT matched all of the images using the EPI tools and an independent observer assessed 31% of the matched images. The lateral correlation coefficient achieved between the 2 RTs was 0.70 and the longitudinal correlation coefficient was 0.76. The coefficient has a positive correlation and validates the reliability of the results between the 2 RTs. The lateral deviation in the matching program includes both right-left and anterior-posterior movement, while the longitudinal deviation indicates superior-inferior movement. No clinically significant differences were found between the groups studied for the rotation results. The results, shown in Figs. 1 and 2, indicate that both the longitudinal and lateral deviations follow a normal distribution, which is similar to the distribution seen by Hunt et al.10 As can be seen in Table 1, the field placement accuracy range of deviations for the polycarbonate system was larger than that of the thermoplastic system. The means were very similar, as were the median and standard deviation. The least-squares mean assessed the absolute mean deviation from the planned isocenter. As can be seen, the thermoplastic system mean deviation was slightly less than the polycarbonate system. Statistical analysis of the data has shown no significant difference in accuracy between the polycarbonate and thermoplastic systems.
Fig. 2. Longitudinal deviation from the isocenter for thermoplastic and polycarbonate treatments.
41
Fig. 3. A comparison of the time taken to produce a thermoplastic mask to a polycarbonate mask.
Production costs Raw material costs. The raw material costs of both systems were obtained from the manufacturers. The polycarbonate material, together with the plaster and plaster bandages were cheaper to purchase than the thermoplastic material. For example, the raw material cost for a small polycarbonate sheet was one sixth the cost of a thermoplastic sheet of similar size and the cost of a large polycarbonate sheet was one seventh the cost of its equivalent thermoplastic sheet. Labor cost. The labor cost of the polycarbonate material was greater due to the labor-intensive process of forming the mask. Figure 3 is a comparison of the time taken to produce 15 polycarbonate and 15 thermoplastic masks. The average labor time to produce a polycarbonate mask was 126 minutes, while the average labor time to produce a thermoplastic mask was 13.6 minutes. The time taken to produce a thermoplastic mask at the WBRC was faster than Gerber et al.11 who quoted 20 to 30 minutes. Unit production cost. The UPC for the polycarbonate mask was $AUD51.32 for the small size and $AUD57.17 for the large size. The UPC for a thermoplastic mask was $AUD48.88 for the small size if used once or $AUD12.22 when used 4 times. A large mask was $AUD74.88 if used once or $AUD18.72 when used 4 times. The above results indicated that the thermoplastic small mask is approximately one quarter the cost of the equivalent polycarbonate mask, while a large thermoplastic mask is approximately one third the cost of its equivalent polycarbonate mask when used 4 times. The thermoplastic system, when re-used, represents a dramatic cost saving to a department. User acceptability of the thermoplastic vs. polycarbonate systems Surveys were distributed between July and August 1997 to 22 members of the planning and treatment RT staff. Thirteen surveys were returned (59%). Surveys were then redistributed between July and August 1999 to
42
Medical Dosimetry
Volume 28, Number 1, 2003
Table 1. Results of field placement accuracy using the Varian electronic portal imaging system Deviation Number of fields matched Range of deviations (mm) Mean (mm) Median (mm) Standard Deviation (mm) Least-squares mean Least-squares Standard Error
Thermoplastic Lateral
Polycarbonate Lateral
Thermoplastic Longitudinal
Polycarbonate Longitudinal
317.0 ⫺7.3–8.1 ⫺0.2 ⫺0.2 1.7 1.51 0.16
335.0 ⫺5.8–9.3 ⫺0.3 ⫺0.1 2.1 1.56 0.18
317.0 ⫺6.6–7.1 ⫺0.3 ⫺0.1 1.8 1.48 0.14
335.0 ⫺7.4–10 0.1 0.2 1.9 1.52 0.16
22 of the then-current planning and treatment RT staff. Seventeen surveys were returned (77%). The results of both surveys are shown in Table 2. The results of the initial survey indicated that RTs at the WBRC preferred the polycarbonate system. The main disadvantage of the thermoplastic system was its opacity. Anatomical landmarks such as tragal notch are routinely matched on the mask to the skin for mask fit. This procedure was difficult using the thermoplastic mask, and RTs perceived the lack of visual verification a distinct disadvantage. The repeat survey indicated a change in attitude toward the thermoplastic system. The main perceived disadvantages of the opacity of the thermoplastic mask was still a slight disadvantage. However, 1 advantage of the thermoplastic system, recognized as important by the RTs, was the benefit to the patient in requiring only 1 visit for mask production immediately presimulation. Also, the ease of construction of a thermoplastic mask reduced the production time to approximately one tenth that of a polycarbonate mask. DISCUSSION AND CONCLUSIONS When assessing field placement, radically-treated patients were compared with those receiving palliative treatments. While it may be thought the 2 groups are insignificantly different, it should be noted that the latter have underlying conditions that tend to make them more difficult to treat. These include swelling, pain, discomfort, and mobility problems. These factors often make them more difficult to stabilize. Despite the bias in the
palliative group of patients, the thermoplastic system achieved a comparable accuracy to the polycarbonate system. The results of this study are difficult to compare with other studies because of different methods of analysis. The results for the polycarbonate system masks can be compared with another field placement accuracy study that used EPIs.6 The mean deviation achieved by Bel et al.6 was ⫺0.8 mm laterally and ⫺0.8 mm longitudinally compared with the results of ⫺0.2 mm to ⫺0.1 mm achieved in this study. The standard deviation achieved by Bel et al.,6 although clinically comparable, was more accurate with a value of 1.5 mm laterally and 1.6 mm longitudinally compared to the WBRC results of 2.1 mm laterally and 1.9 mm longitudinally for polycarbonate masks. The intertreatment standard deviation achieved by Fiorino et al. for thermoplastic masks was less than 2 mm,16 which is comparable with the WBRC results. The absolute deviation from the isocenter was 2.38 mm for the thermoplastic system and 2.46 mm for the polycarbonate system and is comparable to the range cited by other studies1,11,12,14 and more complicated devices.13,17 When assessing which thermoplastic system to trial as an alternative to the polycarbonate system, a few alternate thermoplastic materials were chosen to determine which 1 was more suited to the department. The system that was chosen for its combination of malleability, reusability, and cost was the Biomet Efficast system. Fixed costs associated with each system were not con-
Table 2. Results of the initial and repeat surveys on RT staff opinion Category Ease of construction Time taken to construct Achieving required position Marking of cast In simulator Accuracy Viewing of anatomy Fit of cast Ease of adjustment Stability of cast Field adjustment
Initial Preference
Follow-up Preference
Opinion changed
Thermoplastic Thermoplastic Polycarbonate
Thermoplastic Thermoplastic Either
No No Yes
Polycarbonate Polycarbonate Polycarbonate Polycarbonate Either Either Either Either
Polycarbonate Polycarbonate Either Polycarbonate Thermoplastic Thermoplastic Either Either
No No Yes No Yes Yes No No
Comparison of two immobilization systems ● L. LORD et al.
sidered, as the analysis dealt with UPCs and were similar. Fixed cost items included baseboards, standard neck shapes, and vacuum bags. The main cost advantage of the thermoplastic system is its ability to be used at least 4 times. The WBRC is currently using the thermoplastic system for all patients requiring an immobilization shell. There are 2 groups of patients—larynx and electron treatment areas—that routinely have areas cut out of the mask. If thermoplastic masks have areas cut out, they lose their elastic properties and are unable to be reused. A mask on its third or fourth application is more cost effective to use for these types of treatment. Once used, the masks are washed in warm soapy water and dried flat, ready to be used for the next patient. If the mask has bodily fluids from the previous patient, it is discarded. When the thermoplastic system was initially introduced to the WBRC, user acceptance was low. There has been a gradual acceptance from the RTs by initially restricting use of the thermoplastic masks for palliative treatments. The main disadvantage of this system to our RTs is the masks’ opacity. After 2 years of routine clinical use and hands-on experience, the improved acceptance can be summed up by 1 survey comment stating “my attitude to the thermoplastic system has improved dramatically with increased use.” The recommendation to obtain full acceptance of the thermoplastic system would be to develop a translucent mask material. This study indicates that the initial perception of the RTs that the polycarbonate system is more accurate has changed. This change in attitude has been aided by the knowledge that the accuracy of the systems is comparable. There is no statistically significant difference between the 2 systems in terms of treatment accuracy. This study has shown the thermoplastic system to be more cost effective even when it is reused at least twice, although it can be used up to 4 times. The thermoplastic system has the additional advantage for the patient that it requires 1 less visit to the department and reduces the time dedicated to forming a mask by the staff. The initial recommendation was to use this thermoplastic system for all patients, excluding those patients whose masks are routinely cut out unless the mask material is on its final use. Four years after the thermoplastic masks have been introduced, they are now being used exclusively for all brain and head-and-neck patients.
43
Acknowledgments—The authors thank Kay Hatherly and Professor Alan Rodger for their editing skills. The authors also thank BioMet for their financial support in assisting to present this paper at the NZMRT conference.
REFERENCES 1. Niewald, M.; Lehmann, W.; Uhlmann, U.; et al. Plastic material used to optimise radiotherapy of the head and neck tumours and the mammary carcinoma. Radiother. Oncol 11:55–63; 1988. 2. Bentel, G.; Marks, L.; Hendren, K.; et al. Comparison of two head and neck immobilisation systems. Int. J. Radiat. Oncol. Biol. Phys 38:867–73; 1997. 3. Bentel, G.; Marks, L.; Sherouse, G.; et al. A customised head and neck support system. Int. J. Radiat. Oncol. Biol. Phys 32:245–8; 1995. 4. Rabinowitz, I.; Broomberg, J.; Goitein, M.; et al. Accuracy of radiation field alignment in clinical practice. Int. J. Radiat. Oncol. Biol. Phys 11:1857–67; 1985. 5. Weltens, C.; Kesteloot, K.; Vandevelde, G.; et al. Comparison of plastic and Orfit® masks for patient head fixation during radiotherapy: Precision and costs. Int. J. Radiat. Oncol. Biol. Phys 33:499 –507; 1995. 6. Bel, A.; Keus, R.; Vijlbrief, R.; et al. Set up deviations in wedged pair irradiation of parotid gland and tonsillar tumours measured with an electronic portal imaging device. Radiother. Oncol 37: 153–9; 1995. 7. Marsh, R.; Balter, J.; Evans, V.; et al. Design and analysis of an immobilisation and repositioning system for treatment of neck malignancies. Med. Dosim 22:293–7; 1997. 8. Sweeney, R.; Bale, R.; Vogele, M.; et al. Repositioning accuracy: Comparison of a non-invasive head holder with thermoplastic mask for fractionated radiotherapy and a case report. Int. J. Radiat. Oncol. Biol. Phys 41:475–83; 1998. 9. Hess, C.; Kortmann, R.; Jany, R.; et al. Accuracy of field alignment in radiotherapy of head and neck cancer utilising individualised face mask immobilisation: A retrospective analysis of clinical practise. Radiother. Oncol 34:69 –72; 1995. 10. Hunt, M.; Kutcher, G.; Burman, C.; et al. The effect of setup uncertainties on the treatment of nasopharynx cancer. Int. J. Radiat. Oncol. Biol. Phys 27:437–47; 1993. 11. Gerber, R.; Marks, J.; Purdy, J. The use of thermal plastics for immobilisation of treatments during radiotherapy. Int. J. Radiat. Oncol. Biol. Phys 8:1461–2; 1982. 12. Huizenga, H.; Levendag, P.; De Porre, M.; et al. Accuracy in radiation field alignment in head and neck cancer: A prospective study. Radiother. Oncol 11:181–7; 1988. 13. Thornton, A.; Ten Haken, R.; Gerhardsson, A.; et al. Threedimensional motion analysis of an improved head immobilisation system for simulation, CT, MRI and PET imaging. Radiother. Oncol 20:224 –8; 1991. 14. Dunscombe, P.; Fox, K.; Loose, S.; et al. The investigation and rectification of field placement errors in the delivery of complex head and neck fields. Int. J. Radiat. Oncol. Biol. Phys 26:155–61; 1993. 15. SAS Institute Inc., Carey, NC. 16. Fiorino, C.; Cattaneo, G.M.; Del Vecchio, A.; et al. Skin-sparing reduction effects of thermoplastics used for immobilisation in head and neck radiotherapy. Radiother. Oncol 30:267–70; 1994. 17. Verhey, L.; Goitein, M.; McNulty, P.; et al. Precise positioning of treatments for radiation therapy. Int. J. Radiat. Oncol. Biol. Phys 8:289 –94; 1982.