- Email: [email protected]
Graduated Block Technique for the Treatment of Paranasal Sinus Tumors
Medrcal Ihsrmerrv. Vol. 16. pp. 199-204 Pnnted I” the U.S.A. All rights reserved.
l
0739-021 l/91 $3.00 + .ocl Copyright 0 1991 American Assoaat~on of MedIcal Dosimetrists
Third Place Smithers/AAMD 1991 Writing Competition GRADUATED
BLOCK TECHNIQUE FOR THE TREATMENT OF PARANASAL SINUS TUMORS
MATTToBLER,CMD,JANALYNPROWS,M.D.,M.S., ~~~DE~sD.LEAvITT,PH.D. University of Utah Health Sciences Center, Division of Radiation Oncology, 50 N. Medical Drive, Rm. BOSO,Salt Lake City, UT 84132 Abstract-Cancers of the head and neck often present difficult dosimetric challenges; tumors of the paranasal sinuses, often advanced at diagnosis, pose several problems in treatment planning. Adequate coverage of involved areas often necessitates inclusion of the ipsilateral orbit due to direct extension of disease; sparing the uninvolved contralateral orbit may be difficult, especially if the superior nasal cavity and ethmoid sinus must be treated. We will report on a technique that allows delivery of a relatively homogeneous dose to a treatment volume that includes the paranasal sinus and ipsilateral orbit, with significant sparing of the anterior chamber of the contralateral eye. This technique uses a heavily weighted anterior field designed to deliver 100% to a plane posterior to the lens of each eye. From this plane posteriorly, lateral wedged fields are employed to increase the dose as the anterior contribution decreases. To achieve maximum homogeneity would require a wedge angle of greater than 60”, the maximum wedge commonly available. To overcome this restraint, this technique uses multiple lateral wedged fields whose anterior field edges graduate in a posterior direction allowing for further compensation of the anterior field’s fall-off. Film densitometry using a Rando phantom* is used to verify the technique. Key Words: Sinus, Orbit, Blocking.
INTRODUCTION
and upper nasal cavity, with significant sparing of the contralateral orbit and anterior chamber of the contralateral eye. This technique combines a heavily weighted anterior field with opposed wedged lateral fields modified by graduating blocks. This technique will be compared with two commonly used techniques: the first consists of a wedged pair, and the second is a three-field technique comprising wedged laterals and an anterior.2
Over one-half of all cancer patients will receive radiotherapy as a component of care in the management of their disease. Numerous technical advances in the past decade, including an increase in the use of computerized tomography and magnetic resonance imaging to assist in treatment planning and the expanding role of computerized dosimetry, have given dosimet&s and radiation oncologists the ability to execute sophisticated treatment plans; this allows optimization of dosimetry and a potential decrease in the morbidity of therapy. Cancers of the head and neck often present difficult dosimetric challenges; tumors of the paranasal sinuses are often advanced at the time of diagnosis and often require radiotherapy fields that by necessity must include the ipsilateral orbit in order to adequately cover all areas involved. Sparing of the uninvolved contralateral eye may be difficult to accomplish, especially if the ethmoid sinuses must be treated. In addition, it is often difficult to achieve a homogeneous dose distribution anteriorly when the ethmoid sinuses, which may extend anteriorly nearly to the inner canthus,’ and/or upper nasal cavity are involved. A treatment technique has been designed that allows delivery of a relatively homogeneous dose to a treatment volume that includes the paranasal sinuses and the ipsilateral orbit, as well as the ethmoids
METHODS
AND MATERIALS
Positioning Simulation was performed on a Rando head phantom for the purpose of treatment planning and verification. A head position was chosen that placed the anterior facial surface in a plane horizontal to the beam incidence of an anterior field. An isocenter was placed in the center of the treatment volume in both the left-to-right and superior-to-inferior dimensions for the anterior planning film. For the lateral planning film, a center was chosen that would cover the treatment volume posteriorly, but would place the anterior field edge approximately at the level of the bony canthus. The exact location ofthe isocenter and necessary field edges for each evaluated treatment plan was determined at the time the computer treatment planning was done. Surface markings that would be used for isocenter location reproducibility were placed on the phantom. Prior to obtaining a needed treatment planning CT scan, sinus had to be created in the
* Alderson Research Laboratory, Stamford, CT. 199
200
Medical Dosimetry
Volume 16, Number 4, 1991
phantom based on its bony anatomy to more closely approximate the true anatomy of an actual patient. Transverse treatment planning CT scans were obtained through the treatment volume with a one centimeter spacing between images, assuring that one scan was located at the level of the orbits (Fig. 1). The scans were taken with the phantom in treatment position and radio-opaque markers were placed on the lateral and anterior skin surface in a superior to inferior direction corresponding to the horizontal beam axis to aid in the orientation of the scans to the planned position. From these scans, sinus density measurements were also determined. Tumor volumes were drawn by the physician on the CT scans and simulation films. Blocking was also drawn by the physician on the anterior and lateral films to aid in the treatment planning process. Any oblique fields that might be planned would need a second simulation to acquire the necessary films for field shaping. The multiple patient contours with their corresponding internal structures as determined from the CT scans were entered into the treatment planning computer. Treatment planning Three treatment techniques were evaluated on the treatment planning system. Plan 1 consisted of an anterior wedged field with a beam weighting of 47% at the isocenter and a right anterior wedged oblique field with a weighting of 53% at the isocenter. The hinge angle between these two
Fig. 2. Plan I is a wedged pair consisting of an anterior field weighted to deliver 47% at the isocenter and a right anterior oblique weighted to deliver 53% at the isocenter. This plan delivers a high dose of 103% and the treatment volume is covered by the 92% isodose line. Dose to the uninvolved orbit is greater than desired to reduce the probability of cataract formation.
Fig. 1. Representation of the transverse CT scan located at the level of the orbits. Internal structures and tumor volume have been outlined.
beams was 50”. Both fields were designed using 60” wedges and 4 Mev photons (Fig. 2). Plan 2 utilized a heavily weighted anterior field with two wedged lateral fields. Both of the laterals had fixed anterior borders entering at a point just posterior to the lens of each eye. The anterior field was designed with a 15’ wedge, and 60” wedges were used for each lateral field. A weighting of 70% from the anterior, 20% from the left lateral, and 10% from the right lateral was used. 4 Mev photons were used in the planning of all three fields (Fig. 3). Plan 3 is the graduated block technique. This technique has been used in the treatment planning of many of our paranasal sinus patients. Planning for the use of this technique for treatment of the phantom was started with the design of an anterior field. A wedge was used in the field to compensate for the obliquity of the patient’s anterior surface. This anterior wedged field was weighted in a way that delivered a dose of 100% to a plane posterior to the lens of each eye, approximately at the level of the mid-orbit (Fig. 4). Posterior to this 100% isodose plane the dose delivered from the anterior field began to decrease. Lateral wedged fields were needed in order to maintain dose homogeneity throughout the treatment volume. The
Treatment of paranasal sinus tumors 0 M. TOBLER etal.
201
Fig. 3. Plan 2 is a 3-field technique consisting of a wedged anterior field weighted to deliver 70% at the isocenter, a left lateral with a weighted to deliver 20% at the isocenter, and a right lateral with a weighted to deliver 10% at the isocenter. This plan delivers a high dose of 108%, and the treatment volume anteriorly is covered by the 90% isodose line. The 10% isodose line encompassed less of the uninvolved orbit than the previous plan and is more posterior to the lens; however. dose delivered to the lens is still of concern.
first wedged lateral field was designed so that its anterior geometric edge was placed at the level of the 100% isodose line created by the anterior field. This first wedged lateral was weighted in a way that helped to move the 100% isodose line in a posterior direction without introducing a significant high dose region (Fig. 5). The resultant isodose distribution was again evaluated and the second lateral field was designed. Most linear accelerators come equipped with a 60” wedge as the maximum thickness available. This is true of the equipment available at our institution. The possibility for dose uniformity would be improved with a wedge angle of greater thickness. Because such a wedge was not readily available, it was necessary for further adjustment of the dose homogeneity to be accomplished by other methods. A block was designed for this second opposing lateral wedged field that placed its anterior geometric edge at the posterior edge of the newly noted 100% isodose line. Multiple wedged lateral fields were necessary, in this case four, each with an anterior field edge designed more posterior than the previously planned laterals (Fig. 6). This field arrangement was necessary to achieve the desired volume coverage and uniformity. With the addition of each lateral field it was necessary to evaluate the dose variation and determine the need for adjustment of the exact beam weighting and location of the anterior border to achieve the desired distribution (Fig. 7). Again 4 Mev photons were used for all fields.
Fig. 4. Anterior only isodose distribution showing the placement of the 100% isodose line at the desired depth and the resultant high and low dose regions.
Fig. 5. The addition of a left lateral field with its anterior geometric edge placed at the level of the 100% isodose line created by the anterior field helps to move the newly created 100% isodose line posteriorly to begin better volume coverage.
Medical Dosimetry
202
Volume 16, Number 4. I99 1
4 ::::::::: ..........:::............... .:::: ............,........ ...::: ..F.y.’ ....... iii;iIiiii::::::::::::::::: .............-3 ....:: ..........................:::::: ..............:::::::::::: .........*.:: ii*.:::::i:>.: ........................................
..x~iii~: .::::::::v.i:i::::: ................... ‘:ift!g~: :::::iB::::::::.. ::::*.::::i’:z ‘iigj iiii::::::::ii:::: .;::::::::::+:::: :iZE’:; ::::::i::i:>.i.::: ‘5 ;iiiiit$z.risr: :::. ::::I:.::::::::::: ii :::: . . . . . . t.:::::: ~~~~ii:i~~~ ii .:::.....:::::::::. ...:::::::: ...... ii .$iti.i:q::.‘:: ...:::::......:::. .:: ::: sii~:~.:$.;;y:; .ixi m::::::::: ...._. .::r iii.::;:::::::::... :::: ..........::::: .:::: y:y:iiizi:ii+:::: .:iii:.i .~~~.~.~.~~~~:, ...........::::ii ,............... :i:.i:;i’:i: ~
3
5 .iiiiiiiiiiiiiiiii:iiiaiiiiiiiiiiiiiiiiii~::::: ........................................~::::.
........................ ...................... :::::::::::::::::::::::::::::::::::::::::::::::: ::::::::: ::::::: ....:::::::::::::::::::::::::::: :::::::::::::::::::::::::::: ::::::::::::::. iiiiiiiiliiiiiiiiiiiiitiiiii ::::::::::::. ::::::::::. :::::::::. :::::::::::::::::::::::::::. ::::::::::::::::::::::::::: :::::::. ::::::::::::::::::::::::::: ::::::. ::::::::::::::::::::::::::: ::::: ::::::::::::::::::::::::::: :::: ::::::::::::::::::::::::::: ::: ::::::::::::::::::::::::::: ii ::::::::::::::::::::::::::: ::::::::::::::::::::::::::: ii ::::::::::::::::::::::::::: ::::::::::::::::::::::::::: .ii !iiiiitiiiitiiiil:lililiiii ::: .iiiiiiiiiiililiiiiiitiiiiii ::: ::: :::::::::::::::::::::::::::: :::: :::::::::::::::::::::::::::: :::: ::z:::::::::::::::::::::::: ::::: :::::. .:::::::::::::::::::::::::::: ........................... :::::::::::::::::::::::::::: ::::::. :::::::::::::::::::::::::::: ::::::::.. .::::::::::::.......:::::::::::::::::::::::::::: . . . . . . . . . . . . ............. ...... .. . ~ Fig. 6. Blocking used for the anterior field and the graduated blocking used for each of the four lateral fields. (1) The anterior field is designed and delivers a dose of 100% to a plane posterior to the lens of both eyes. (2) The anterior edge ofthe first lateral field is placed at the level of the 100% isodose line created by the anterior field. (3) The design of a second lateral with the anterior edge graduated slightly posteriorly helps in the movement toward the desired volume coverage. (4) Each lateral field is designed with the anterior edge slightly more posterior then the one previously designed. (5) The design ofthe final lateral block increases the posterior dose so that the entire volume is covered by the 100% isodose line.
Verification Because Plan 1 and Plan 2 are similar to already established and widely used treatment techniques, no verification was done. Blocks for the graduated block technique were manufactured based on the blocking that had been both drawn by the physician and designed by the dosimet& to produce the best dosimetric results with the planned field arrangement. The blocks were checked against the simulator film after construction to verify their positioning. Kodak Ready pack XV2t verification film was cut into a square that would just fit inside the phantom’s anterior, posterior, and lateral surfaces, and the cut ends were taped with black photographic tape. The films were then placed between the sections of the head phantom and secured with an adjustable wood clamp. The phantom was aligned to the markings that had been placed at the time of simulation. The planned treatment was performed on a Varian Clinac 4/ 100* using monitor units that should * Varian Associates, Inc., Palo Alto, CA. + Eastman Kodak Company, Rochester, NY.
deliver a total dose of 100 cGy to the 100% isodose line. At that time, control films were also exposed using the same ready pack film. These exposures were made using tissue equivalent polystyrene sheets located at a distance of 100 cm from the source to the surface. Each film was placed at a depth of 1.0 cm below the surface, which is d,, for the treatment machine used, and exposed. Exposures were made at doses of 0, 20, 40, 60, 80, 100, and 120 cGy. Film densitometry, based on the values determined by the control films, was performed by tracking dose values for use in comparison with the computer predictions. Agreement between the computer predictions and the densitometry measurements was noted to be within approximately +2%. RESULTS
Plan I This technique (Fig. 2) shows good coverage of the anterior treatment volume and structures anteriorly. Posteriorly, the inhomogeneous dose distribution across the volume does not appear to be ade-
Treatment of paranasal sinus tumors 0 M. TOBLER et al.
203
quate. A high dose of 103% and a low dose of 92% covering the posterior tumor margin result in a dose variation across the treatment volume of 11%. If this patient had disease posterior and lateral to the already involved orbit, the posterior edge of the oblique field would need to be increased. This would result in an increased dose to the temporal lobe of the involved side. If the treatment were delivered as planned, sparing of the uninvolved orbital lens would be minimal. Should the patient’s position change slightly during the treatment period, dose to this opposite eye could increase substantially. Plan 2 When a three-field technique (Fig. 3) is used, some problems may be overcome. Disease that may occur posterior and lateral to the involved orbit can be more easily treated without a great increase in the posterior temporal lobe. Placing the anterior border of both lateral fields directly posterior to the lens of the uninvolved eye helps to minimize the resultant dose; however, significant dose is still delivered with little margin given for the possibility of undesired patient positioning changes. The necessary placement of these lateral fields also places an area of decreased dose in the anterior treatment volume. With a high dose of 108% and a low dose of 92%, the variation across the treatment volume is 16%. Plan 3 With the use of the graduated block technique (Fig. 7) many of the previously mentioned problems are resolved. As with the three-field technique, posterior and lateral volume coverage can be accomplished without an extreme dose to the temporal lobe. Further evaluation shows the same 108% dose excess as noted in Plan 2, with complete target volume coverage by the 100% isodose line resulting in a variation of only 8%. The design of the lateral fields posterior to the middle of the uninvolved orbit greatly decreases dose to the contralateral lens and also allows for greater sparing should an inaccuracy in setup occur. In order to achieve these results, the dose to the posterior brain has been increased slightly due to the heavy weighting of the anterior field.
DISCUSSION We believe the first two techniques to be acceptable and adequate in many situations. There are, however, situations that require improvement of anterior, posterior, and lateral volume coverage and greater uninvolved orbital sparing. With the heavy anterior field weighting of the graduated block technique, adequate coverage of the required volume is more easily achieved without a significant increase in dose to nor-
Fig. 7. A composite isodose distribution of the graduated block technique allows for more adequate volume coverage and better orbital sparing. This plan delivers a high dose of 108% and the treatment volume is covered by the 100% isodose line. Dose to the lens of the uninvolved orbit has been reduced to approximately 2%.
mal tissue. This is due to the freedom allowed the dosimetrist to place the 100% isodose line from the anterior field at the desired depth posterior to the orbital lens and to design the lateral fields accordingly. If, as the clinician evaluates the resultant dose distribution, the anterior high dose region is deemed unacceptable, adjustment of the anterior beam weight and subsequent movement of the anterior border of the lateral field are possible. This could also decrease exist dose to the posterior structures, but the desired greater sparing of the uninvolved orbit might be forfeited. When compared with these accepted treatment techniques, the graduated block technique allows the dose to the uninvolved orbit to be decreased due to the ability to position the lateral field more posteriorly. This greater uninvolved orbital sparing is particularly desirable because of the possibility of some patient movement during treatment in spite of immobilization efforts. This is desirable because it allows for the possibility of undesired patient movement. In designing the lateral fields, the anterior edge of the first lateral field should be collimated, if possible, so that transmission through blocking to the lens of the uninvolved eye is avoided. The same collimator setting should be used for the design of subsequent laterals, with adjustment of the anterior border accomplished by the design of blocking. In newer therapy units, which allow independent asymmetric motion of the collimator jaws, additional reduction in dose to the shielded regions can be achieved through defining the
204
Medical Dosimetry
anterior field edge using the independent collimator jaw. In evaluating each plane, an appropriate head position chosen at the time of simulation would produce an anterior block design that would be in a horizontal direction. This is helpful because a block designed in the initial plane will usually be correct for the other planes being evaluated. If a different head position was chosen, the evaluation of each plane would create a block in a diagonal position, similar to the positioning of the anterior facial plane. Multiple planes must be calculated in order to evaluate the dose homogeneity and volume coverage. In other planes, areas of high or low dose can usually be corrected by modifying the design of one or more of the graduated block levels for that plane.
Volume 16, Number 4, 1991
CONCLUSION In some unique treatment planning situations the graduated block technique improves dose coverage of the treatment volume and reduces the dose delivered to the lens of the contralateral orbit, thereby reducing the likelihood of cataract formation. This technique is in routine use at our institution. REFERENCES I. Million, R.R.; Cassisi, N.J.; Hamlin, D.J. Nasal vestibule, nasal cavity, and paranasal sinuses. In: Million, R.R.; Cassisi, N.J., editors. Management of head and neck cancer: A multidisciplinary approach. Philadelphia: J.B. Lippincott; 1984: 407-444. 2. Wang, C.C. Radiation therapy for head and neck neoplasms: Indications, techniques, and results. Chicago: Year Book Medical Publishers, Inc.; 1990: 289-293.
l
0739-021 l/91 $3.00 + .ocl Copyright 0 1991 American Assoaat~on of MedIcal Dosimetrists
Third Place Smithers/AAMD 1991 Writing Competition GRADUATED
BLOCK TECHNIQUE FOR THE TREATMENT OF PARANASAL SINUS TUMORS
MATTToBLER,CMD,JANALYNPROWS,M.D.,M.S., ~~~DE~sD.LEAvITT,PH.D. University of Utah Health Sciences Center, Division of Radiation Oncology, 50 N. Medical Drive, Rm. BOSO,Salt Lake City, UT 84132 Abstract-Cancers of the head and neck often present difficult dosimetric challenges; tumors of the paranasal sinuses, often advanced at diagnosis, pose several problems in treatment planning. Adequate coverage of involved areas often necessitates inclusion of the ipsilateral orbit due to direct extension of disease; sparing the uninvolved contralateral orbit may be difficult, especially if the superior nasal cavity and ethmoid sinus must be treated. We will report on a technique that allows delivery of a relatively homogeneous dose to a treatment volume that includes the paranasal sinus and ipsilateral orbit, with significant sparing of the anterior chamber of the contralateral eye. This technique uses a heavily weighted anterior field designed to deliver 100% to a plane posterior to the lens of each eye. From this plane posteriorly, lateral wedged fields are employed to increase the dose as the anterior contribution decreases. To achieve maximum homogeneity would require a wedge angle of greater than 60”, the maximum wedge commonly available. To overcome this restraint, this technique uses multiple lateral wedged fields whose anterior field edges graduate in a posterior direction allowing for further compensation of the anterior field’s fall-off. Film densitometry using a Rando phantom* is used to verify the technique. Key Words: Sinus, Orbit, Blocking.
INTRODUCTION
and upper nasal cavity, with significant sparing of the contralateral orbit and anterior chamber of the contralateral eye. This technique combines a heavily weighted anterior field with opposed wedged lateral fields modified by graduating blocks. This technique will be compared with two commonly used techniques: the first consists of a wedged pair, and the second is a three-field technique comprising wedged laterals and an anterior.2
Over one-half of all cancer patients will receive radiotherapy as a component of care in the management of their disease. Numerous technical advances in the past decade, including an increase in the use of computerized tomography and magnetic resonance imaging to assist in treatment planning and the expanding role of computerized dosimetry, have given dosimet&s and radiation oncologists the ability to execute sophisticated treatment plans; this allows optimization of dosimetry and a potential decrease in the morbidity of therapy. Cancers of the head and neck often present difficult dosimetric challenges; tumors of the paranasal sinuses are often advanced at the time of diagnosis and often require radiotherapy fields that by necessity must include the ipsilateral orbit in order to adequately cover all areas involved. Sparing of the uninvolved contralateral eye may be difficult to accomplish, especially if the ethmoid sinuses must be treated. In addition, it is often difficult to achieve a homogeneous dose distribution anteriorly when the ethmoid sinuses, which may extend anteriorly nearly to the inner canthus,’ and/or upper nasal cavity are involved. A treatment technique has been designed that allows delivery of a relatively homogeneous dose to a treatment volume that includes the paranasal sinuses and the ipsilateral orbit, as well as the ethmoids
METHODS
AND MATERIALS
Positioning Simulation was performed on a Rando head phantom for the purpose of treatment planning and verification. A head position was chosen that placed the anterior facial surface in a plane horizontal to the beam incidence of an anterior field. An isocenter was placed in the center of the treatment volume in both the left-to-right and superior-to-inferior dimensions for the anterior planning film. For the lateral planning film, a center was chosen that would cover the treatment volume posteriorly, but would place the anterior field edge approximately at the level of the bony canthus. The exact location ofthe isocenter and necessary field edges for each evaluated treatment plan was determined at the time the computer treatment planning was done. Surface markings that would be used for isocenter location reproducibility were placed on the phantom. Prior to obtaining a needed treatment planning CT scan, sinus had to be created in the
* Alderson Research Laboratory, Stamford, CT. 199
200
Medical Dosimetry
Volume 16, Number 4, 1991
phantom based on its bony anatomy to more closely approximate the true anatomy of an actual patient. Transverse treatment planning CT scans were obtained through the treatment volume with a one centimeter spacing between images, assuring that one scan was located at the level of the orbits (Fig. 1). The scans were taken with the phantom in treatment position and radio-opaque markers were placed on the lateral and anterior skin surface in a superior to inferior direction corresponding to the horizontal beam axis to aid in the orientation of the scans to the planned position. From these scans, sinus density measurements were also determined. Tumor volumes were drawn by the physician on the CT scans and simulation films. Blocking was also drawn by the physician on the anterior and lateral films to aid in the treatment planning process. Any oblique fields that might be planned would need a second simulation to acquire the necessary films for field shaping. The multiple patient contours with their corresponding internal structures as determined from the CT scans were entered into the treatment planning computer. Treatment planning Three treatment techniques were evaluated on the treatment planning system. Plan 1 consisted of an anterior wedged field with a beam weighting of 47% at the isocenter and a right anterior wedged oblique field with a weighting of 53% at the isocenter. The hinge angle between these two
Fig. 2. Plan I is a wedged pair consisting of an anterior field weighted to deliver 47% at the isocenter and a right anterior oblique weighted to deliver 53% at the isocenter. This plan delivers a high dose of 103% and the treatment volume is covered by the 92% isodose line. Dose to the uninvolved orbit is greater than desired to reduce the probability of cataract formation.
Fig. 1. Representation of the transverse CT scan located at the level of the orbits. Internal structures and tumor volume have been outlined.
beams was 50”. Both fields were designed using 60” wedges and 4 Mev photons (Fig. 2). Plan 2 utilized a heavily weighted anterior field with two wedged lateral fields. Both of the laterals had fixed anterior borders entering at a point just posterior to the lens of each eye. The anterior field was designed with a 15’ wedge, and 60” wedges were used for each lateral field. A weighting of 70% from the anterior, 20% from the left lateral, and 10% from the right lateral was used. 4 Mev photons were used in the planning of all three fields (Fig. 3). Plan 3 is the graduated block technique. This technique has been used in the treatment planning of many of our paranasal sinus patients. Planning for the use of this technique for treatment of the phantom was started with the design of an anterior field. A wedge was used in the field to compensate for the obliquity of the patient’s anterior surface. This anterior wedged field was weighted in a way that delivered a dose of 100% to a plane posterior to the lens of each eye, approximately at the level of the mid-orbit (Fig. 4). Posterior to this 100% isodose plane the dose delivered from the anterior field began to decrease. Lateral wedged fields were needed in order to maintain dose homogeneity throughout the treatment volume. The
Treatment of paranasal sinus tumors 0 M. TOBLER etal.
201
Fig. 3. Plan 2 is a 3-field technique consisting of a wedged anterior field weighted to deliver 70% at the isocenter, a left lateral with a weighted to deliver 20% at the isocenter, and a right lateral with a weighted to deliver 10% at the isocenter. This plan delivers a high dose of 108%, and the treatment volume anteriorly is covered by the 90% isodose line. The 10% isodose line encompassed less of the uninvolved orbit than the previous plan and is more posterior to the lens; however. dose delivered to the lens is still of concern.
first wedged lateral field was designed so that its anterior geometric edge was placed at the level of the 100% isodose line created by the anterior field. This first wedged lateral was weighted in a way that helped to move the 100% isodose line in a posterior direction without introducing a significant high dose region (Fig. 5). The resultant isodose distribution was again evaluated and the second lateral field was designed. Most linear accelerators come equipped with a 60” wedge as the maximum thickness available. This is true of the equipment available at our institution. The possibility for dose uniformity would be improved with a wedge angle of greater thickness. Because such a wedge was not readily available, it was necessary for further adjustment of the dose homogeneity to be accomplished by other methods. A block was designed for this second opposing lateral wedged field that placed its anterior geometric edge at the posterior edge of the newly noted 100% isodose line. Multiple wedged lateral fields were necessary, in this case four, each with an anterior field edge designed more posterior than the previously planned laterals (Fig. 6). This field arrangement was necessary to achieve the desired volume coverage and uniformity. With the addition of each lateral field it was necessary to evaluate the dose variation and determine the need for adjustment of the exact beam weighting and location of the anterior border to achieve the desired distribution (Fig. 7). Again 4 Mev photons were used for all fields.
Fig. 4. Anterior only isodose distribution showing the placement of the 100% isodose line at the desired depth and the resultant high and low dose regions.
Fig. 5. The addition of a left lateral field with its anterior geometric edge placed at the level of the 100% isodose line created by the anterior field helps to move the newly created 100% isodose line posteriorly to begin better volume coverage.
Medical Dosimetry
202
Volume 16, Number 4. I99 1
4 ::::::::: ..........:::............... .:::: ............,........ ...::: ..F.y.’ ....... iii;iIiiii::::::::::::::::: .............-3 ....:: ..........................:::::: ..............:::::::::::: .........*.:: ii*.:::::i:>.: ........................................
..x~iii~: .::::::::v.i:i::::: ................... ‘:ift!g~: :::::iB::::::::.. ::::*.::::i’:z ‘iigj iiii::::::::ii:::: .;::::::::::+:::: :iZE’:; ::::::i::i:>.i.::: ‘5 ;iiiiit$z.risr: :::. ::::I:.::::::::::: ii :::: . . . . . . t.:::::: ~~~~ii:i~~~ ii .:::.....:::::::::. ...:::::::: ...... ii .$iti.i:q::.‘:: ...:::::......:::. .:: ::: sii~:~.:$.;;y:; .ixi m::::::::: ...._. .::r iii.::;:::::::::... :::: ..........::::: .:::: y:y:iiizi:ii+:::: .:iii:.i .~~~.~.~.~~~~:, ...........::::ii ,............... :i:.i:;i’:i: ~
3
5 .iiiiiiiiiiiiiiiii:iiiaiiiiiiiiiiiiiiiiii~::::: ........................................~::::.
........................ ...................... :::::::::::::::::::::::::::::::::::::::::::::::: ::::::::: ::::::: ....:::::::::::::::::::::::::::: :::::::::::::::::::::::::::: ::::::::::::::. iiiiiiiiliiiiiiiiiiiiitiiiii ::::::::::::. ::::::::::. :::::::::. :::::::::::::::::::::::::::. ::::::::::::::::::::::::::: :::::::. ::::::::::::::::::::::::::: ::::::. ::::::::::::::::::::::::::: ::::: ::::::::::::::::::::::::::: :::: ::::::::::::::::::::::::::: ::: ::::::::::::::::::::::::::: ii ::::::::::::::::::::::::::: ::::::::::::::::::::::::::: ii ::::::::::::::::::::::::::: ::::::::::::::::::::::::::: .ii !iiiiitiiiitiiiil:lililiiii ::: .iiiiiiiiiiililiiiiiitiiiiii ::: ::: :::::::::::::::::::::::::::: :::: :::::::::::::::::::::::::::: :::: ::z:::::::::::::::::::::::: ::::: :::::. .:::::::::::::::::::::::::::: ........................... :::::::::::::::::::::::::::: ::::::. :::::::::::::::::::::::::::: ::::::::.. .::::::::::::.......:::::::::::::::::::::::::::: . . . . . . . . . . . . ............. ...... .. . ~ Fig. 6. Blocking used for the anterior field and the graduated blocking used for each of the four lateral fields. (1) The anterior field is designed and delivers a dose of 100% to a plane posterior to the lens of both eyes. (2) The anterior edge ofthe first lateral field is placed at the level of the 100% isodose line created by the anterior field. (3) The design of a second lateral with the anterior edge graduated slightly posteriorly helps in the movement toward the desired volume coverage. (4) Each lateral field is designed with the anterior edge slightly more posterior then the one previously designed. (5) The design ofthe final lateral block increases the posterior dose so that the entire volume is covered by the 100% isodose line.
Verification Because Plan 1 and Plan 2 are similar to already established and widely used treatment techniques, no verification was done. Blocks for the graduated block technique were manufactured based on the blocking that had been both drawn by the physician and designed by the dosimet& to produce the best dosimetric results with the planned field arrangement. The blocks were checked against the simulator film after construction to verify their positioning. Kodak Ready pack XV2t verification film was cut into a square that would just fit inside the phantom’s anterior, posterior, and lateral surfaces, and the cut ends were taped with black photographic tape. The films were then placed between the sections of the head phantom and secured with an adjustable wood clamp. The phantom was aligned to the markings that had been placed at the time of simulation. The planned treatment was performed on a Varian Clinac 4/ 100* using monitor units that should * Varian Associates, Inc., Palo Alto, CA. + Eastman Kodak Company, Rochester, NY.
deliver a total dose of 100 cGy to the 100% isodose line. At that time, control films were also exposed using the same ready pack film. These exposures were made using tissue equivalent polystyrene sheets located at a distance of 100 cm from the source to the surface. Each film was placed at a depth of 1.0 cm below the surface, which is d,, for the treatment machine used, and exposed. Exposures were made at doses of 0, 20, 40, 60, 80, 100, and 120 cGy. Film densitometry, based on the values determined by the control films, was performed by tracking dose values for use in comparison with the computer predictions. Agreement between the computer predictions and the densitometry measurements was noted to be within approximately +2%. RESULTS
Plan I This technique (Fig. 2) shows good coverage of the anterior treatment volume and structures anteriorly. Posteriorly, the inhomogeneous dose distribution across the volume does not appear to be ade-
Treatment of paranasal sinus tumors 0 M. TOBLER et al.
203
quate. A high dose of 103% and a low dose of 92% covering the posterior tumor margin result in a dose variation across the treatment volume of 11%. If this patient had disease posterior and lateral to the already involved orbit, the posterior edge of the oblique field would need to be increased. This would result in an increased dose to the temporal lobe of the involved side. If the treatment were delivered as planned, sparing of the uninvolved orbital lens would be minimal. Should the patient’s position change slightly during the treatment period, dose to this opposite eye could increase substantially. Plan 2 When a three-field technique (Fig. 3) is used, some problems may be overcome. Disease that may occur posterior and lateral to the involved orbit can be more easily treated without a great increase in the posterior temporal lobe. Placing the anterior border of both lateral fields directly posterior to the lens of the uninvolved eye helps to minimize the resultant dose; however, significant dose is still delivered with little margin given for the possibility of undesired patient positioning changes. The necessary placement of these lateral fields also places an area of decreased dose in the anterior treatment volume. With a high dose of 108% and a low dose of 92%, the variation across the treatment volume is 16%. Plan 3 With the use of the graduated block technique (Fig. 7) many of the previously mentioned problems are resolved. As with the three-field technique, posterior and lateral volume coverage can be accomplished without an extreme dose to the temporal lobe. Further evaluation shows the same 108% dose excess as noted in Plan 2, with complete target volume coverage by the 100% isodose line resulting in a variation of only 8%. The design of the lateral fields posterior to the middle of the uninvolved orbit greatly decreases dose to the contralateral lens and also allows for greater sparing should an inaccuracy in setup occur. In order to achieve these results, the dose to the posterior brain has been increased slightly due to the heavy weighting of the anterior field.
DISCUSSION We believe the first two techniques to be acceptable and adequate in many situations. There are, however, situations that require improvement of anterior, posterior, and lateral volume coverage and greater uninvolved orbital sparing. With the heavy anterior field weighting of the graduated block technique, adequate coverage of the required volume is more easily achieved without a significant increase in dose to nor-
Fig. 7. A composite isodose distribution of the graduated block technique allows for more adequate volume coverage and better orbital sparing. This plan delivers a high dose of 108% and the treatment volume is covered by the 100% isodose line. Dose to the lens of the uninvolved orbit has been reduced to approximately 2%.
mal tissue. This is due to the freedom allowed the dosimetrist to place the 100% isodose line from the anterior field at the desired depth posterior to the orbital lens and to design the lateral fields accordingly. If, as the clinician evaluates the resultant dose distribution, the anterior high dose region is deemed unacceptable, adjustment of the anterior beam weight and subsequent movement of the anterior border of the lateral field are possible. This could also decrease exist dose to the posterior structures, but the desired greater sparing of the uninvolved orbit might be forfeited. When compared with these accepted treatment techniques, the graduated block technique allows the dose to the uninvolved orbit to be decreased due to the ability to position the lateral field more posteriorly. This greater uninvolved orbital sparing is particularly desirable because of the possibility of some patient movement during treatment in spite of immobilization efforts. This is desirable because it allows for the possibility of undesired patient movement. In designing the lateral fields, the anterior edge of the first lateral field should be collimated, if possible, so that transmission through blocking to the lens of the uninvolved eye is avoided. The same collimator setting should be used for the design of subsequent laterals, with adjustment of the anterior border accomplished by the design of blocking. In newer therapy units, which allow independent asymmetric motion of the collimator jaws, additional reduction in dose to the shielded regions can be achieved through defining the
204
Medical Dosimetry
anterior field edge using the independent collimator jaw. In evaluating each plane, an appropriate head position chosen at the time of simulation would produce an anterior block design that would be in a horizontal direction. This is helpful because a block designed in the initial plane will usually be correct for the other planes being evaluated. If a different head position was chosen, the evaluation of each plane would create a block in a diagonal position, similar to the positioning of the anterior facial plane. Multiple planes must be calculated in order to evaluate the dose homogeneity and volume coverage. In other planes, areas of high or low dose can usually be corrected by modifying the design of one or more of the graduated block levels for that plane.
Volume 16, Number 4, 1991
CONCLUSION In some unique treatment planning situations the graduated block technique improves dose coverage of the treatment volume and reduces the dose delivered to the lens of the contralateral orbit, thereby reducing the likelihood of cataract formation. This technique is in routine use at our institution. REFERENCES I. Million, R.R.; Cassisi, N.J.; Hamlin, D.J. Nasal vestibule, nasal cavity, and paranasal sinuses. In: Million, R.R.; Cassisi, N.J., editors. Management of head and neck cancer: A multidisciplinary approach. Philadelphia: J.B. Lippincott; 1984: 407-444. 2. Wang, C.C. Radiation therapy for head and neck neoplasms: Indications, techniques, and results. Chicago: Year Book Medical Publishers, Inc.; 1990: 289-293.