1111.J. Radrntmn Oncology Biol Phys, Vol. 19, pp. 183-188 Printed in the U.S.A. All rights reserved.
Copyright
0360-3016/90 $3.00 + 40 0 1990 Pergamon Press plc
??Technical Innovations and Notes
A TECHNIQUE FOR TREATING LOCAL BREAST CANCER USING SINGLE SET-UP POINT AND ASYMMETRIC COLLIMATION ULF F. ROSENOW,
PH.D.,
EDWARD S. VALENTINE,
M.D.
A
AND LAWRENCE W. DAVIS, M.D.
Department of Radiation Oncology, Albert Einstein College of Medicine and Montefiore Medical Center, 111 East 210th St., NY, NY 10467 Using both pairs of asymmetric jaws of a linear accelerator local-regional breast cancer may be treated from a single set-up point. This point is placed at the abutment of the supraclavicuhu fields with the medial and lateral tangential fields. Positioning the jaws to create a half-beam superiorly permits treatment of the supraclavicular field. Positioning both jaws asymmetrically at midline to define a single beam in the inferoanterior quadrant permits treatment of the breast from medial and lateral tangents. The highest possible matching accuracy between the supraclavicular and tangential fields is inherently provided by this technique. For treatment of all fields at 100 cm source to axis distance (SAD) the lateral placement and depth of the set-up point may be determined by simulation and simple trigonometry. We elaborate on the clinical procedure. For the technologists treatment of all fields from a single set-up point is simple and efficient. Since the tissue at the superior border of the tangential fields is generally firmer than in mid-breast, greater accuracy in day-to-day set-up is permitted. This technique eliminates the need for table angles even when tangential fields only are planned. Because of half-beam collimation the limit to the tangential field length is 20 cm. Means will be suggested to overcome this limitation in the few cases where it occurs. Another modification is suggested for linear accelerators with only one independent pair of jaws. Breast cancer, Irradiation technique, Asymmetric fields, Single set-up point.
mits treatment of the breast from medial and lateral tangents. We describe the technique in this paper.
INTRODUCTION
of early breast cancer after conservative surgery requires treating a supraclavicular field in addition to medial and lateral tangential fields, coplanarity at the abutting borders is necessary to minimize dose inhomogeneities. Other considerations include minimizing the dose to the ipsilateral and contralateral lungs, and with treatment of the left breast, minimizing the dose to the heart. Finally, with the increasing number of patients undergoing conservative surgery plus radiation, the set-up for breast irradiation should be simple for the technologists to localize, simulate, and perform. It should also be accurate in its reproduction from treatment to treatment. Using both pairs of asymmetric jaws of a linear accelerator* local-regional breast cancer may be treated from a single isocentric set-up point at the junction of the tangential and supraclavicular fields. Positioning the inferior jaw at the center to create a non-divergent beam permits treatment of the supraclavicular field. Positioning both the superior jaw and the posterior jaw at the midline to create non-divergent inferoanterior quadrant beams perWhen irradiation
METHODS
AND
MATERIALS
The new treatment technique is illustrated in Figures l-3. The positioning ofthe isocenter in the matching plane in connection with half-beam collimation guarantees an exact and reproducible matching. The final position of the isocenter in the patient results from the intention to irradiate the breast with symmetrically positioned tangential fields (Fig. 2) and will consequently make the medial and lateral field borders of the supraclavicular field asymmetric in respect to the isocenter. From the beam’seye view of the tangential breast fields in Figure 3 it can be seen that by means of half-beam collimation in both dimensions only the inferoanterior quadrants of the corresponding symmetric fields are used. Whenever the longitudinal axes of the tangential fields are not horizontal a small tangential field block is needed to provide exact matching. These field blocks are derived from the colli-
Reprint requests to: Edward Scott Valentine, M.D. acknowledgement is made to Mr. V. Sivapalasingam for his technical assistance and to Joseph Ting, Ph.D. for his review of the manuscript.
Accepted for publication 24 January 1990. * Philips SL-25.
Acknowledgements-Grateful
183
184
1. J. Radiation Oncology 0 Biology 0 Physics
July 1990, Volume 19, Number 1
mid-breast plane
lateral
tangential
Fig. 1. Schematic representation of the set-up of the supraclavicular and the tangential fields using a single set-up point (3). 1, 2, and 3 represent the positions the isocenter is moved to during localization.
mator angle at the superior border of the tangential field, as described below under localization and simulation. Localization and simulation The patient is positioned supine on the simulator table with the ipsilateral arm abducted 90” or greater. The arm is supported in a reproducible way by a specifically designed arm board. The medial border is marked 1 to 2
cm beyond the palpable extent of medial breast tissue. The lateral border is marked 1 to 2 cm beyond the palpable extent of lateral breast tissue. The inferior border is marked 1 to 2 cm below the inferior extent of palpable breast tissue. The superior border is marked at the level of the sternoclavicular joint, or, if no supraclavicular field is planned, 1 to 2 cm superior to the superior extent of palpable breast tissue. These marks must be final because
Fig. 2. Symmetric tangential field arrangement in the mid-breast plane (see Fig. 1) with the pseudo-isocenter (2 in Fig. 1) at half separation. The tangentials are half-collimated posteriorly for coplanarity. It is seen how y, s, and x are derived.
Breast:
singleset-uppoint0
U. F. ROSENOW el al.
185
hall collimated
Fig. 3. Beam’s eye view of the tangential fields which are half-collimated in both field axes and matched to the supraclav by means of the small tangential field block. This would also represent the fluoroscopic view after adjusting the table height, resulting in the final position of the isocenter and the set-up point SSD (3 in Fig. 1). all future calculations in this technique depend upon them; that is, any correction of the marker positions which might result from fluoroscopy will require returning to this point. Medial and lateral border separation is measured. For an approximate measure of the gantry angle a device for measuring the angle of incline between medial and lateral borders may be used at this time. To distinguish one border from the other under radiography radioopaque metal arrows are stuck to the medial and lateral borders. The medial and lateral arrow are applied facing superiorly and inferiorly, respectively. A preliminary beam light field is set up so that the central axis bisects the medial border. The table is raised or lowered until the skin surface is at 100 SAD. This creates a half-beam in the anterior-posterior dimension, and establishes a center on the skin surface that will not stray as the gantry is swung. The gantry angle is determined by swinging the gantry until the medial and lateral arrows superimpose under fluoroscopy. The collimator angle is determined by rotating the collimator until the posterior half beam is roughly parallel with the general slope of the chest wall underneath the breast. The next step is to align the beam with the mid-separation between the medial and the lateral borders. The lateral distance, X, to be moved is calculated by the following formula: X = s/2 X sin gamma where s is the medial-lateral separation and gamma is the Gantry Angle. The unit for both X and s is cm.
Before actual movement the gantry is swung back to vertical. For accuracy the lateral distance is read off a lateral scale at the foot of the simulator table. Immediately following the lateral table movement by X the table top is moved longitudinally to let the central axis of the beam align with the junction or matching plane of the tangential and supraclavicular fields; this is the set-up point. To determine the depth the gantry is swung to the medial gantry angle and the collimator rotated to the medial collimator angle. Under fluoroscopy the simulation table is raised until the medial and lateral arrows line up with the central hairline between the anterior and posterior half-beams. There are now half-beams in both the anterior-posterior and superior-inferior dimensions. The center of the beam under fluoroscopy is in fact the isocenter at 100 SAD. The anterior border is adjusted until it clears the apex of the breast by 1.5 to 2.0 cm and the inferior border until it clears the caudal most aspect of the breast by 1.5 to 2.0 cm. The set-up source to skin distance (SSD) is determined by swinging the gantry back up to the vertical and reading the SSD off the optical distance indicator. Medial and lateral simulation films are taken. To eliminate divergence of the angled tangential fields into the supraclavicular field, blocks are drawn on the tangential simulation films by dropping a vertical line from the anterior tangential border to intersect the isocenter. The angle of these blocks will equal to the collimator angle. Even when a supraclavicular field is not planned these blocks will eliminate divergence of the tangential fields into the
186
I. J. Radiation Oncology 0 Biology 0 Physics
Fig. 4. Three-field of the set-up.
set-up of an individual
July 1990, Volume 19, Number I
patient plan to the film/slab
phantom
for the purpose of quality control
patient’s ipsilateral arm. The tangential simulation is now complete. For simulation of the supraclavicular field the gantry is swung to 15” off the vertical to avoid the spinal cord. The jaws are closed to correspond to the medial, lateral, and superior borders as marked on the patient. Any blocks needed are drawn on the simulation film. RESULTS
The accuracy of matching beams was investigated in a polystyrene slab phantom with film densitometry (Fig. 4). The film was irradiated by pairs of equal beams offset to geometrically match. Independent of field size the best matching was achieved if an additional offset of 1 mm was applied to one of the fields. This value obviously results from the specific form of the penumbra generated by the pair of jaws in question and also from the servocontrol mechanism of field size and field offset. Optimal offsets would vary from machine to machine. We considered offsets in units of full millimeters since field size and offset are stated only in full millimeters on this machine*. This result was highly reproducible and was also found valid for patient treatment plans (Fig. 5, center) as set up on the phantom in Figure 4. In consequence, we routinely offset the supraclavicular field by an additional 1 mm which is simply included in the field definition as stored
-
A’ -
Fig. 5. Densitometric demonstration of quality of offset field matching. Left: Three film exposures in slab phantom of match-. ing vertical fields with 0, 1, and 2 mm additional offset (from top to bottom). Middle: Quality control film as produced according to Figure 4 for a real patient plan with supraclavicular field (top rectangle with humoral block) and two tangentials (with exactly matching superior, S, and posterior, P, edges). The supraclavicular field has additional offsets of 0, 1, and 2 mm (from top to bottom). The graph on the right shows densitometric curves obtained along lines A - A’ for A) two matched offset fields of 10 X 20 cm2, B) two such fields of 2.5 X 5 cm*, and C) the patient plan. Notice the same matching behavior in all cases.
Breast: single set-up point 0 U. F. ROSENOWet al.
in the computer. The set-up of asymmetric (or offset) beams is largely facilitated by an assisted set-up option of the computerized console which automatically sets the beam parameters. A quality control procedure for individual treatment plans is routinely applied. For this purpose the film/polystyrene slab phantom, as shown in Figure 4, is used prior to the first treatment. The isocenter is set to the film plane and the film is exposed to the three beams set up consecutively by the assisted set-up from stored patient data, except that the monitor units are adjusted to give the desired film density. This procedure allows for a check of the matching accuracy including the correct positioning of the tangential field blocks and the posterior tangential beam border coplanarity. Although the clinical procedures are simplified, safer, and more accurate with this method, the dosimetry is more complicated. The dosimetry of offset open beams has been dealt with by Khan et al. (3) whose findings we have confirmed. Our treatment planning system can handle offset open fields, although the monitor units produced usually need manual corrections of less than 3%. For prescription dose distributions are routinely calculated in both the mid-breast plane (Fig. 1) and in an irregular field calculation at 3 cm depth from a central point in the supraclavicular field. In the matching plane, however, the treatment planning system has not been found sufficiently reliable to produce a meaningful dose distribution. Consequently, we rely on the inherent accuracy of matching provided by the technique, as demonstrated in Figure 5. DISCUSSION In the three field technique for definitive irradiation of early breast cancer after conservative surgery major dose inhomogeneities can occur at the border between the tangents and the supraclavicular field. Divergence of the tangents into the lungs can be avoided by angling the central axes of the tangential beams to greater than 180 degrees so that the posterior field borders are coplanar ( 1,2,4, 5, 6, 7, 8). Similarly, divergence of the supraclavicular field into the lung can be eliminated with a caudal half-beam block (1,2,4,5,6,7,8). The greatest challenge, however, is to provide a smooth dose transition between the supraclavicular and tangential beams. In 1980, Svensson et al. devised a hanging block which, combined with angulation of the treatment table and a protractor designed to maintain accuracy of the junction between the supraclavicular and tangentials, sharply reduced the dose inhomogeneities (8). This technique was modified with the introduction of rotating half beam blocks (8) and tangential field comer blocks (6). However, it requires specially designed hardware and set-up can be complicated. Podgorsak et al. devised a monoisocentric set-up for all three fields which makes use of a universally designed rotating half-beam block which is used superiorly for the supraclavicular field, and inferiorly for the medial and lateral
187
tangential fields (5). The isocenter is set at the skin surface of the junction between tangential and supraclavicular fields and centrally in the lateral supraclavicular dimension; this results in an asymmetric positioning of the tangential fields at mid-breast. The advantages of this treatment are: (a) a single set-up point for all three fields SO that the patient does not have to be set up once for the tangentials and a second time for the supraclavicular, (b) consequent accuracy in placement of fields without the need for a protractor, (c) elimination of table angulation, and (d) minimization of dose to lung (5). Recently, Conte et al. have modified this monoisocentric technique with the use of individualized blocks. The tangential beams are kept horizontal relative to the axis of gantry rotation. In the superior half of the portal a half-beam block eliminates superior divergence of the tangentials into the supraclavicular field. In the inferior half of the portal a posterior block is cut to shield the lung just below the chest wall. In fact, both these blocks are cast into one large one. Because the posterior blocking is not half-beam the tangentials must be offset in gantry angle to make the posterior beams coplanar as in the Podgorsak et al. technique (5). Although the authors (5) report that individualized blocks avoid the excess weight of standard half-beam blocks, they go on to say that the typical weight of the tangential blocks is 10 to 15 kg. They caution that these blocks permit approximately 5% transmission of the delivered dose to the target volume (2). The technique presented in this paper takes advantage of the asymmetric jaws of the linear accelerator. There are several advantages to this technique in addition to those demonstrated by Podgorsak et al. (5) Since both sets of jaws are asymmetric half-beams can be created in two dimensions. This eliminates the problem of the higher transmission through half-beam blocks as compared to collimators. Once the gantry angle is determined the posterior borders of the tangentials are perforce coplanar. However, one may want to purposely deviate from coplanarity of the posterior tangential field borders to reduce the hot spots near the medial and lateral aspects of the breast caused by lung transmission. Also, the tangentials are kept in a symmetric arrangement at mid-breast, indicated by the pseudo-isocenter at half separation (Fig. 2). With our technique no entry and exit points of the beam central axis need to be marked on the breast. Location of a single isocenter at the inferior aspect of the supraclavicular field places the set-up point on skin which is less mobile than at the center or sides of the breast, thus ensuring greater accuracy of set-up. An extension of the monoisocentric technique to include an additional axillary field or an additional posterior supraclavicular field is possible with the same ease and matching accuracy by setting them up to the same isocenter and using half-beam and asymmetric collimation. An additional advantage of a single set-up point is the ease of quality control with a simple film/slab phantom
188
1. J. Radiation Oncology 0 Biology 0 Physics
and exposure of the film to the three fields set up with the assisted set-up feature of the computerized console (Figs. 4 and 5). In addition, the exact position of the tangential blocks is checked prior to the first set-up and repeatedly with consecutive set-ups. The critical block shadow edge must coincide with the lateral laser beams and not necessarily with the marking of the matching plane on the patient’s skin which may vary slightly from day to day. The localization is simple and requires only one trigonometric calculation. Set-up is easy: the patient is triangulated, the beam is centered on the set-up point with gantry angle set to zero, and table set to proper SSD for 100 SAD isocentric technique. The three fields are called in consecutively at the computerized console and set up using the assisted set-up option for automatic set-up of the beam parameters. With each field the appropriate block tray also has to be inserted. No angulation, raising, or lowering of the table is needed. Although patients whose axillary dissections show no positive lymph nodes do not generally receive supraclavicular radiation this technique is still useful for its ease and accuracy of set-up and for prevention of beam divergence into the ipsilateral arm. If an internal mammary field is indicated it may be drawn initially with the subsequent medial-lateral separation being determined from the medial tangential-internal mammary junction. The monoisocentric simulation proceeds as described. There are two disadvantages to this technique. The first is that maximum field length at 100 SAD is 40 cm; therefore the maximum half-beam length for the tangentials is 20 cm. For tangential breast fields longer than 20 cm two alternative approaches exist: (a) more of the superior breast aspect could be incorporated into the supraclavicular field as long as the increased volume of lung irradiated is acceptable, or (b) the tangentials could be extended superiorly beyond the isocenter and made coplanar to the matching plane by an additional slight table rotation. The isocenter of the supraclavicular field would have to be shifted superiorly accordingly. Approach (b) can obviously no longer benefit from the inherent simplicity and accuracy of the monoisocentric set-up, as can approach (a). The second disadvantage of this technique is that wedges
July 1990, Volume 19, Number I
covering both halves of the beam length are not available. We are having two brass wedges custom made to fit offset field sizes up to 40 cm in field length and 20 cm in the wedge direction (Ting, J., personal communication). The present method could still be used with linear accelerators having a single set of asymmetric jaws provided the following modifications are applied. A half-beam block needs to be inserted to shield the inferior half of the supraclavicular field. The present tangential blocks need to be enlarged to cover the part of the tangential fields superior to the matching plane. The independent jaws would be used (a) to set the medial and lateral supraclavicular field limits asymmetrically and (b) to half-collimate the tangential fields posteriorly. We feel that the simplification of the clinical procedures involved in this monoisocentric new technique, supported also by the assisted set-up, and the shift of more complex aspects of it to the areas of dosimetry and treatment planning are steps in the right direction. CONCLUSIONS In a linear accelerator with two pairs of independent jaws half-beams in two dimensions may be created. Taking advantage of this capability we have developed a monoisocentric technique for the three field treatment of early breast cancer which avoids the use of heavy blocks and the higher transmissions that they permit as compared to the collimator transmission. The technique is inherently easy and accurate, allows for the symmetric irradiation of the breast, and for easy quality control prior to the first treatment. The localization and simulation are equally simple while dosimetry and treatment planning are more sophisticated. Posterior supraclavicular or axillary beams, if intended, can easily be added with the same technique and same improved matching behavior. The technique can be adjusted to include an additional internal mammary field, be it a photon or an electron field. A modification is described to deal with tangential fields requiring a field length larger than 20 cm, the maximum for halfcollimated fields. We have successfully applied the principle of the present monoisocentric technique to other applications where field matching appeared to be favorable.
REFERENCES 1. Casebow,
2.
3. 4.
5.
M. P. Matching of adjacent radiation beams for isocentric radiotherapy. Br. J. Radiol. 57:735-740; 1984. Conte, G.; Nascimben, 0.; Turcato, G.; Police, R.; Idi, M. B.; Belleri, L. M.; Bergoglio, F.; Simonato, F.; Langranco, S.; Bugin, F.; Bortot, N. Three-field isocentric technique for breast irradiation using individualized shielding blocks. Int. J. Radiat. Oncol. Biol. Phys. 14: 1299-1305; 1988. Khan, F. M.; Gerbi, B. J.; Deibel, F. C. Dosimetry of asymmetric x-ray collimators. Med. Phys. 13:936-94 1; 1986. Lichter, A. S.; Fraass, B. A.; van de Geijn, J.; Padikal, T. N. A technique for field matching in primary breast irradiation. Int. J. Radiat. Oncol. Biol. Phys. 9:263-270; 1983. Podgorsak, E. B.; Gosselin, M.; Pla, M.; Kim, T. H.; Freeman, C. R. A simple isocentric technique for irradiation of
the breast chest wall and peripheral lymphatics. 57:57-63; 1984.
Br. J. Radiol.
Siddon, R. L.; Buck, B. A.; Harris, J. R.; Svensson, G. K. Three-field technique for breast irradiation using tangential field corner blocks. Int. J. Radiat. Oncol. Biol. Phys. 9:583588; 1983. Siddon, R. L.; Tonnesen, G. L.; Svensson, G. K. Threefield technique for breast treatment using a rotatable halfbeam block. Int. J. Radiat. Oncol. Biol. Phys. 7:1473-1477; 1981. Svensson, G. K.; Bjarngard, B. E.; Larsen, R. D.; Levene, M. B. A modified three-field technique for breast treatment. Int. J. Radiat. Oncol. Biol. Phys. 6:689-694; 1980.
Copyright
0360-3016/90 $3.00 + 40 0 1990 Pergamon Press plc
??Technical Innovations and Notes
A TECHNIQUE FOR TREATING LOCAL BREAST CANCER USING SINGLE SET-UP POINT AND ASYMMETRIC COLLIMATION ULF F. ROSENOW,
PH.D.,
EDWARD S. VALENTINE,
M.D.
A
AND LAWRENCE W. DAVIS, M.D.
Department of Radiation Oncology, Albert Einstein College of Medicine and Montefiore Medical Center, 111 East 210th St., NY, NY 10467 Using both pairs of asymmetric jaws of a linear accelerator local-regional breast cancer may be treated from a single set-up point. This point is placed at the abutment of the supraclavicuhu fields with the medial and lateral tangential fields. Positioning the jaws to create a half-beam superiorly permits treatment of the supraclavicular field. Positioning both jaws asymmetrically at midline to define a single beam in the inferoanterior quadrant permits treatment of the breast from medial and lateral tangents. The highest possible matching accuracy between the supraclavicular and tangential fields is inherently provided by this technique. For treatment of all fields at 100 cm source to axis distance (SAD) the lateral placement and depth of the set-up point may be determined by simulation and simple trigonometry. We elaborate on the clinical procedure. For the technologists treatment of all fields from a single set-up point is simple and efficient. Since the tissue at the superior border of the tangential fields is generally firmer than in mid-breast, greater accuracy in day-to-day set-up is permitted. This technique eliminates the need for table angles even when tangential fields only are planned. Because of half-beam collimation the limit to the tangential field length is 20 cm. Means will be suggested to overcome this limitation in the few cases where it occurs. Another modification is suggested for linear accelerators with only one independent pair of jaws. Breast cancer, Irradiation technique, Asymmetric fields, Single set-up point.
mits treatment of the breast from medial and lateral tangents. We describe the technique in this paper.
INTRODUCTION
of early breast cancer after conservative surgery requires treating a supraclavicular field in addition to medial and lateral tangential fields, coplanarity at the abutting borders is necessary to minimize dose inhomogeneities. Other considerations include minimizing the dose to the ipsilateral and contralateral lungs, and with treatment of the left breast, minimizing the dose to the heart. Finally, with the increasing number of patients undergoing conservative surgery plus radiation, the set-up for breast irradiation should be simple for the technologists to localize, simulate, and perform. It should also be accurate in its reproduction from treatment to treatment. Using both pairs of asymmetric jaws of a linear accelerator* local-regional breast cancer may be treated from a single isocentric set-up point at the junction of the tangential and supraclavicular fields. Positioning the inferior jaw at the center to create a non-divergent beam permits treatment of the supraclavicular field. Positioning both the superior jaw and the posterior jaw at the midline to create non-divergent inferoanterior quadrant beams perWhen irradiation
METHODS
AND
MATERIALS
The new treatment technique is illustrated in Figures l-3. The positioning ofthe isocenter in the matching plane in connection with half-beam collimation guarantees an exact and reproducible matching. The final position of the isocenter in the patient results from the intention to irradiate the breast with symmetrically positioned tangential fields (Fig. 2) and will consequently make the medial and lateral field borders of the supraclavicular field asymmetric in respect to the isocenter. From the beam’seye view of the tangential breast fields in Figure 3 it can be seen that by means of half-beam collimation in both dimensions only the inferoanterior quadrants of the corresponding symmetric fields are used. Whenever the longitudinal axes of the tangential fields are not horizontal a small tangential field block is needed to provide exact matching. These field blocks are derived from the colli-
Reprint requests to: Edward Scott Valentine, M.D. acknowledgement is made to Mr. V. Sivapalasingam for his technical assistance and to Joseph Ting, Ph.D. for his review of the manuscript.
Accepted for publication 24 January 1990. * Philips SL-25.
Acknowledgements-Grateful
183
184
1. J. Radiation Oncology 0 Biology 0 Physics
July 1990, Volume 19, Number 1
mid-breast plane
lateral
tangential
Fig. 1. Schematic representation of the set-up of the supraclavicular and the tangential fields using a single set-up point (3). 1, 2, and 3 represent the positions the isocenter is moved to during localization.
mator angle at the superior border of the tangential field, as described below under localization and simulation. Localization and simulation The patient is positioned supine on the simulator table with the ipsilateral arm abducted 90” or greater. The arm is supported in a reproducible way by a specifically designed arm board. The medial border is marked 1 to 2
cm beyond the palpable extent of medial breast tissue. The lateral border is marked 1 to 2 cm beyond the palpable extent of lateral breast tissue. The inferior border is marked 1 to 2 cm below the inferior extent of palpable breast tissue. The superior border is marked at the level of the sternoclavicular joint, or, if no supraclavicular field is planned, 1 to 2 cm superior to the superior extent of palpable breast tissue. These marks must be final because
Fig. 2. Symmetric tangential field arrangement in the mid-breast plane (see Fig. 1) with the pseudo-isocenter (2 in Fig. 1) at half separation. The tangentials are half-collimated posteriorly for coplanarity. It is seen how y, s, and x are derived.
Breast:
singleset-uppoint0
U. F. ROSENOW el al.
185
hall collimated
Fig. 3. Beam’s eye view of the tangential fields which are half-collimated in both field axes and matched to the supraclav by means of the small tangential field block. This would also represent the fluoroscopic view after adjusting the table height, resulting in the final position of the isocenter and the set-up point SSD (3 in Fig. 1). all future calculations in this technique depend upon them; that is, any correction of the marker positions which might result from fluoroscopy will require returning to this point. Medial and lateral border separation is measured. For an approximate measure of the gantry angle a device for measuring the angle of incline between medial and lateral borders may be used at this time. To distinguish one border from the other under radiography radioopaque metal arrows are stuck to the medial and lateral borders. The medial and lateral arrow are applied facing superiorly and inferiorly, respectively. A preliminary beam light field is set up so that the central axis bisects the medial border. The table is raised or lowered until the skin surface is at 100 SAD. This creates a half-beam in the anterior-posterior dimension, and establishes a center on the skin surface that will not stray as the gantry is swung. The gantry angle is determined by swinging the gantry until the medial and lateral arrows superimpose under fluoroscopy. The collimator angle is determined by rotating the collimator until the posterior half beam is roughly parallel with the general slope of the chest wall underneath the breast. The next step is to align the beam with the mid-separation between the medial and the lateral borders. The lateral distance, X, to be moved is calculated by the following formula: X = s/2 X sin gamma where s is the medial-lateral separation and gamma is the Gantry Angle. The unit for both X and s is cm.
Before actual movement the gantry is swung back to vertical. For accuracy the lateral distance is read off a lateral scale at the foot of the simulator table. Immediately following the lateral table movement by X the table top is moved longitudinally to let the central axis of the beam align with the junction or matching plane of the tangential and supraclavicular fields; this is the set-up point. To determine the depth the gantry is swung to the medial gantry angle and the collimator rotated to the medial collimator angle. Under fluoroscopy the simulation table is raised until the medial and lateral arrows line up with the central hairline between the anterior and posterior half-beams. There are now half-beams in both the anterior-posterior and superior-inferior dimensions. The center of the beam under fluoroscopy is in fact the isocenter at 100 SAD. The anterior border is adjusted until it clears the apex of the breast by 1.5 to 2.0 cm and the inferior border until it clears the caudal most aspect of the breast by 1.5 to 2.0 cm. The set-up source to skin distance (SSD) is determined by swinging the gantry back up to the vertical and reading the SSD off the optical distance indicator. Medial and lateral simulation films are taken. To eliminate divergence of the angled tangential fields into the supraclavicular field, blocks are drawn on the tangential simulation films by dropping a vertical line from the anterior tangential border to intersect the isocenter. The angle of these blocks will equal to the collimator angle. Even when a supraclavicular field is not planned these blocks will eliminate divergence of the tangential fields into the
186
I. J. Radiation Oncology 0 Biology 0 Physics
Fig. 4. Three-field of the set-up.
set-up of an individual
July 1990, Volume 19, Number I
patient plan to the film/slab
phantom
for the purpose of quality control
patient’s ipsilateral arm. The tangential simulation is now complete. For simulation of the supraclavicular field the gantry is swung to 15” off the vertical to avoid the spinal cord. The jaws are closed to correspond to the medial, lateral, and superior borders as marked on the patient. Any blocks needed are drawn on the simulation film. RESULTS
The accuracy of matching beams was investigated in a polystyrene slab phantom with film densitometry (Fig. 4). The film was irradiated by pairs of equal beams offset to geometrically match. Independent of field size the best matching was achieved if an additional offset of 1 mm was applied to one of the fields. This value obviously results from the specific form of the penumbra generated by the pair of jaws in question and also from the servocontrol mechanism of field size and field offset. Optimal offsets would vary from machine to machine. We considered offsets in units of full millimeters since field size and offset are stated only in full millimeters on this machine*. This result was highly reproducible and was also found valid for patient treatment plans (Fig. 5, center) as set up on the phantom in Figure 4. In consequence, we routinely offset the supraclavicular field by an additional 1 mm which is simply included in the field definition as stored
-
A’ -
Fig. 5. Densitometric demonstration of quality of offset field matching. Left: Three film exposures in slab phantom of match-. ing vertical fields with 0, 1, and 2 mm additional offset (from top to bottom). Middle: Quality control film as produced according to Figure 4 for a real patient plan with supraclavicular field (top rectangle with humoral block) and two tangentials (with exactly matching superior, S, and posterior, P, edges). The supraclavicular field has additional offsets of 0, 1, and 2 mm (from top to bottom). The graph on the right shows densitometric curves obtained along lines A - A’ for A) two matched offset fields of 10 X 20 cm2, B) two such fields of 2.5 X 5 cm*, and C) the patient plan. Notice the same matching behavior in all cases.
Breast: single set-up point 0 U. F. ROSENOWet al.
in the computer. The set-up of asymmetric (or offset) beams is largely facilitated by an assisted set-up option of the computerized console which automatically sets the beam parameters. A quality control procedure for individual treatment plans is routinely applied. For this purpose the film/polystyrene slab phantom, as shown in Figure 4, is used prior to the first treatment. The isocenter is set to the film plane and the film is exposed to the three beams set up consecutively by the assisted set-up from stored patient data, except that the monitor units are adjusted to give the desired film density. This procedure allows for a check of the matching accuracy including the correct positioning of the tangential field blocks and the posterior tangential beam border coplanarity. Although the clinical procedures are simplified, safer, and more accurate with this method, the dosimetry is more complicated. The dosimetry of offset open beams has been dealt with by Khan et al. (3) whose findings we have confirmed. Our treatment planning system can handle offset open fields, although the monitor units produced usually need manual corrections of less than 3%. For prescription dose distributions are routinely calculated in both the mid-breast plane (Fig. 1) and in an irregular field calculation at 3 cm depth from a central point in the supraclavicular field. In the matching plane, however, the treatment planning system has not been found sufficiently reliable to produce a meaningful dose distribution. Consequently, we rely on the inherent accuracy of matching provided by the technique, as demonstrated in Figure 5. DISCUSSION In the three field technique for definitive irradiation of early breast cancer after conservative surgery major dose inhomogeneities can occur at the border between the tangents and the supraclavicular field. Divergence of the tangents into the lungs can be avoided by angling the central axes of the tangential beams to greater than 180 degrees so that the posterior field borders are coplanar ( 1,2,4, 5, 6, 7, 8). Similarly, divergence of the supraclavicular field into the lung can be eliminated with a caudal half-beam block (1,2,4,5,6,7,8). The greatest challenge, however, is to provide a smooth dose transition between the supraclavicular and tangential beams. In 1980, Svensson et al. devised a hanging block which, combined with angulation of the treatment table and a protractor designed to maintain accuracy of the junction between the supraclavicular and tangentials, sharply reduced the dose inhomogeneities (8). This technique was modified with the introduction of rotating half beam blocks (8) and tangential field comer blocks (6). However, it requires specially designed hardware and set-up can be complicated. Podgorsak et al. devised a monoisocentric set-up for all three fields which makes use of a universally designed rotating half-beam block which is used superiorly for the supraclavicular field, and inferiorly for the medial and lateral
187
tangential fields (5). The isocenter is set at the skin surface of the junction between tangential and supraclavicular fields and centrally in the lateral supraclavicular dimension; this results in an asymmetric positioning of the tangential fields at mid-breast. The advantages of this treatment are: (a) a single set-up point for all three fields SO that the patient does not have to be set up once for the tangentials and a second time for the supraclavicular, (b) consequent accuracy in placement of fields without the need for a protractor, (c) elimination of table angulation, and (d) minimization of dose to lung (5). Recently, Conte et al. have modified this monoisocentric technique with the use of individualized blocks. The tangential beams are kept horizontal relative to the axis of gantry rotation. In the superior half of the portal a half-beam block eliminates superior divergence of the tangentials into the supraclavicular field. In the inferior half of the portal a posterior block is cut to shield the lung just below the chest wall. In fact, both these blocks are cast into one large one. Because the posterior blocking is not half-beam the tangentials must be offset in gantry angle to make the posterior beams coplanar as in the Podgorsak et al. technique (5). Although the authors (5) report that individualized blocks avoid the excess weight of standard half-beam blocks, they go on to say that the typical weight of the tangential blocks is 10 to 15 kg. They caution that these blocks permit approximately 5% transmission of the delivered dose to the target volume (2). The technique presented in this paper takes advantage of the asymmetric jaws of the linear accelerator. There are several advantages to this technique in addition to those demonstrated by Podgorsak et al. (5) Since both sets of jaws are asymmetric half-beams can be created in two dimensions. This eliminates the problem of the higher transmission through half-beam blocks as compared to collimators. Once the gantry angle is determined the posterior borders of the tangentials are perforce coplanar. However, one may want to purposely deviate from coplanarity of the posterior tangential field borders to reduce the hot spots near the medial and lateral aspects of the breast caused by lung transmission. Also, the tangentials are kept in a symmetric arrangement at mid-breast, indicated by the pseudo-isocenter at half separation (Fig. 2). With our technique no entry and exit points of the beam central axis need to be marked on the breast. Location of a single isocenter at the inferior aspect of the supraclavicular field places the set-up point on skin which is less mobile than at the center or sides of the breast, thus ensuring greater accuracy of set-up. An extension of the monoisocentric technique to include an additional axillary field or an additional posterior supraclavicular field is possible with the same ease and matching accuracy by setting them up to the same isocenter and using half-beam and asymmetric collimation. An additional advantage of a single set-up point is the ease of quality control with a simple film/slab phantom
188
1. J. Radiation Oncology 0 Biology 0 Physics
and exposure of the film to the three fields set up with the assisted set-up feature of the computerized console (Figs. 4 and 5). In addition, the exact position of the tangential blocks is checked prior to the first set-up and repeatedly with consecutive set-ups. The critical block shadow edge must coincide with the lateral laser beams and not necessarily with the marking of the matching plane on the patient’s skin which may vary slightly from day to day. The localization is simple and requires only one trigonometric calculation. Set-up is easy: the patient is triangulated, the beam is centered on the set-up point with gantry angle set to zero, and table set to proper SSD for 100 SAD isocentric technique. The three fields are called in consecutively at the computerized console and set up using the assisted set-up option for automatic set-up of the beam parameters. With each field the appropriate block tray also has to be inserted. No angulation, raising, or lowering of the table is needed. Although patients whose axillary dissections show no positive lymph nodes do not generally receive supraclavicular radiation this technique is still useful for its ease and accuracy of set-up and for prevention of beam divergence into the ipsilateral arm. If an internal mammary field is indicated it may be drawn initially with the subsequent medial-lateral separation being determined from the medial tangential-internal mammary junction. The monoisocentric simulation proceeds as described. There are two disadvantages to this technique. The first is that maximum field length at 100 SAD is 40 cm; therefore the maximum half-beam length for the tangentials is 20 cm. For tangential breast fields longer than 20 cm two alternative approaches exist: (a) more of the superior breast aspect could be incorporated into the supraclavicular field as long as the increased volume of lung irradiated is acceptable, or (b) the tangentials could be extended superiorly beyond the isocenter and made coplanar to the matching plane by an additional slight table rotation. The isocenter of the supraclavicular field would have to be shifted superiorly accordingly. Approach (b) can obviously no longer benefit from the inherent simplicity and accuracy of the monoisocentric set-up, as can approach (a). The second disadvantage of this technique is that wedges
July 1990, Volume 19, Number I
covering both halves of the beam length are not available. We are having two brass wedges custom made to fit offset field sizes up to 40 cm in field length and 20 cm in the wedge direction (Ting, J., personal communication). The present method could still be used with linear accelerators having a single set of asymmetric jaws provided the following modifications are applied. A half-beam block needs to be inserted to shield the inferior half of the supraclavicular field. The present tangential blocks need to be enlarged to cover the part of the tangential fields superior to the matching plane. The independent jaws would be used (a) to set the medial and lateral supraclavicular field limits asymmetrically and (b) to half-collimate the tangential fields posteriorly. We feel that the simplification of the clinical procedures involved in this monoisocentric new technique, supported also by the assisted set-up, and the shift of more complex aspects of it to the areas of dosimetry and treatment planning are steps in the right direction. CONCLUSIONS In a linear accelerator with two pairs of independent jaws half-beams in two dimensions may be created. Taking advantage of this capability we have developed a monoisocentric technique for the three field treatment of early breast cancer which avoids the use of heavy blocks and the higher transmissions that they permit as compared to the collimator transmission. The technique is inherently easy and accurate, allows for the symmetric irradiation of the breast, and for easy quality control prior to the first treatment. The localization and simulation are equally simple while dosimetry and treatment planning are more sophisticated. Posterior supraclavicular or axillary beams, if intended, can easily be added with the same technique and same improved matching behavior. The technique can be adjusted to include an additional internal mammary field, be it a photon or an electron field. A modification is described to deal with tangential fields requiring a field length larger than 20 cm, the maximum for halfcollimated fields. We have successfully applied the principle of the present monoisocentric technique to other applications where field matching appeared to be favorable.
REFERENCES 1. Casebow,
2.
3. 4.
5.
M. P. Matching of adjacent radiation beams for isocentric radiotherapy. Br. J. Radiol. 57:735-740; 1984. Conte, G.; Nascimben, 0.; Turcato, G.; Police, R.; Idi, M. B.; Belleri, L. M.; Bergoglio, F.; Simonato, F.; Langranco, S.; Bugin, F.; Bortot, N. Three-field isocentric technique for breast irradiation using individualized shielding blocks. Int. J. Radiat. Oncol. Biol. Phys. 14: 1299-1305; 1988. Khan, F. M.; Gerbi, B. J.; Deibel, F. C. Dosimetry of asymmetric x-ray collimators. Med. Phys. 13:936-94 1; 1986. Lichter, A. S.; Fraass, B. A.; van de Geijn, J.; Padikal, T. N. A technique for field matching in primary breast irradiation. Int. J. Radiat. Oncol. Biol. Phys. 9:263-270; 1983. Podgorsak, E. B.; Gosselin, M.; Pla, M.; Kim, T. H.; Freeman, C. R. A simple isocentric technique for irradiation of
the breast chest wall and peripheral lymphatics. 57:57-63; 1984.
Br. J. Radiol.
Siddon, R. L.; Buck, B. A.; Harris, J. R.; Svensson, G. K. Three-field technique for breast irradiation using tangential field corner blocks. Int. J. Radiat. Oncol. Biol. Phys. 9:583588; 1983. Siddon, R. L.; Tonnesen, G. L.; Svensson, G. K. Threefield technique for breast treatment using a rotatable halfbeam block. Int. J. Radiat. Oncol. Biol. Phys. 7:1473-1477; 1981. Svensson, G. K.; Bjarngard, B. E.; Larsen, R. D.; Levene, M. B. A modified three-field technique for breast treatment. Int. J. Radiat. Oncol. Biol. Phys. 6:689-694; 1980.
