- Email: [email protected]
Intensity-modulated photon arc therapy for treatment of pleural mesothelioma
Medical Dosimetry, Vol. 27, No. 4, pp. 255–259, 2002 Copyright © 2002 American Association of Medical Dosimetrists Printed in the USA. All rights reserved 0958-3947/02/$–see front matter
PII: S0958-3947(02)00149-8
INTENSITY-MODULATED PHOTON ARC THERAPY FOR TREATMENT OF PLEURAL MESOTHELIOMA MATT TOBLER, C.M.D., GORDON WATSON, M.D., PH.D., and DENNIS D. LEAVITT, PH.D. University of Utah Health Science Center, Salt Lake City, UT (Accepted 15 April 2001)
Abstract—Radiotherapy plays a key role in the definitive or adjuvant management of patients with mesothelioma of the pleural surface. Many patients are referred for radiation with intact lung following biopsy or subtotal pleurectomy. Delivery of efficacious doses of radiation to the pleural lining while avoiding lung parenchyma toxicity has been a difficult technical challenge. Using opposed photon fields produce doses in lung that result in moderate-to-severe pulmonary toxicity in 100% of patients treated. Combined photon-electron beam treatment, at total doses of 4250 cGy to the pleural surface, results in two-thirds of the lung volume receiving over 2100 cGy. We have developed a technique using intensity-modulated photon arc therapy (IMRT) that significantly improves the dose distribution to the pleural surface with concomitant decrease in dose to lung parenchyma compared to traditional techniques. IMRT treatment of the pleural lining consists of segments of photon arcs that can be intensity modulated with varying beam weights and multileaf positions to produce a more uniform distribution to the pleural surface, while at the same time reducing the overall dose to the lung itself. Computed tomography (CT) simulation is critical for precise identification of target volumes as well as critical normal structures (lung and heart). Rotational arc trajectories and individual leaf positions and weightings are then defined for each CT plane within the patient. This paper will describe the proposed rotational IMRT technique and, using simulated isodose distributions, show the improved potential for sparing of dose to the critical structures of the lung, heart, and spinal cord. © 2002 American Association of Medical Dosimetrists. Key Words:
Treatment planning, Intensity modulation, Arc, Mesothelioma.
the lung, deliver doses of 4250 to the 100% isodose level but block out the central lung volume. Anterior and posterior electron fields deliver the same dose to the untreated pleural lining, which is blocked from the anterior and posterior photon fields and is treated to the 90% isodose value. One difficulty encountered when treating with this technique is that it often requires an overlap between the photon and electron fields, creating an undesired high-dose region in the superficial chest wall at the peripheral edges of the electron fields. Secondarily, selection of the appropriate electron energy becomes critical. Electrons have a more difficult time covering the lining of the lung laterally as it begins to slope away from the beam, suggesting the need for use of a higher energy electron beam. In the mid lung region, the lung rises closer to the chest surface, suggesting that a slightly lower energy electron beam be used. It is possible to resolve this problem, however, by choosing the higher of the electron energies to produce the desired lateral coverage and then add compensation to reduce doses delivered to the mid portion of the lung. The falloff dose from these higher energy electrons, however, again results in added undesired dose delivery to the lung. Consequently, patients have a high incidence of pneumonitis as a major complication of therapy. In this work, we propose the use of a new technique that will allow delivery of a more homogeneous dose to the entire pleural lining while at the same time producing better sparing of the healthy lung volume, as well as the
INTRODUCTION Mesothelioma is a tumor that involves the entire outer surface of the lung (pleural lining). The goal of treating patients with this disease is to deliver curative doses (usually ⱖ 4250 cGy) to the entire pleural surface on the involved side, while at the same time minimizing the dose delivered to the uninvolved tissues of the lung itself and surrounding uninvolved structures. Previous published techniques1,2 include both treatment of the entire involved lung with anterior and posterior photon beams and a combination photon/electron beam approach. Treatment with opposed photon beams that treat the entire lung volume results in 1 of 2 outcomes: (a) excessive dose is delivered to the entire lung, resulting in loss of function; or (b) decrease in the delivery of dose to the pleural lining, resulting in inadequate treatment of the desired volume. Treatment using similar fields and a combination of photons and electrons allows the treatment planner to decrease the dose delivered to the lung by placing a carefully selected electron beam directly incident the lung volume. This should allow better limitation of doses delivered to the healthy uninvolved portion of the lung. In this case, the photon fields, which treat the medial, lateral, superior, and inferior aspects of Reprint requests to: M. Tobler, University of Utah, Department of Radiation Oncology, Health Science Center, 50 North Medical Drive, Salt Lake City, UT 84132. E-mail: matthew.tobler@ hsc.utah.edu 255
256
Medical Dosimetry
Volume 27, Number 4, 2002
uninvolved normal structures, such as the heart and spinal cord. This technique combines the use of dynamic multileaf collimations and dose delivery with multiple narrow-beam rotational field projections to producing a “shell” type isodose distribution that treats the entire pleural surface while maintaining a “cold” central core within the lung volume itself. Isodose distributions will be used to illustrate the potential gains in dose uniformity with this technique compared to the combination photon/ electron technique. METHODS AND MATERIALS When treating with the previously described photon/electron technique, the patient may be positioned with his/her arms by his/her sides because the anterior and posterior photon treatment fields can be easily shaped to exclude treatment of the arms. Due to the rotational nature of our dynamic IMRT technique, however, patient data was obtained and evaluated with the patient’s arms in an upright position. For this comparative patient data set, CT scans were obtained with 0.5-cm spacing extending from 3 cm above the superior extent of the lung to 3 cm below the diaphragm. The CT scans were sent to the RAHD treatment planning computer (RAHD Oncology Products, St. Louis, MO). Structures of interest were outlined, including both lungs, the heart, and the spinal cord. For comparison, we will refer to the photon/electron technique previously described in the literature.1 CT information and treatment planning was used to evaluate (a) the desired electron energies, (b) the potential need for placement of compensation in the electron fields, and (c) the possible need for overlap between the photon and electron fields. The anterior and posterior photon fields were designed to block out the central portion of the lung to a margin of 1 cm inside the projected lung volume in all directions. Blocks were also designed to exclude portions of the liver and stomach. Electron fields were matched to the blocked portion of the photon fields in the lung, and included a slight overlap to ensure adequate coverage of the desired volume. As previously mentioned, with this technique, dose to the pleural surface was limited to the 90% value for the electron fields. For creation of the IMRT technique, the isocenter was placed in the central portion of the lung to allow for rotation of the beam around the entire lung volume. Beams were created at fixed positions every 10° of gantry rotation (Fig. 1). Blocking for each individual beam trajectory was created so that each plane had a 0.5-cm margin both inside and outside the pleural lining of the lung to produce the desired volume coverage (Fig. 2). Each beam projection was independently evaluated through beam’s-eye view to allow for adjustment of the field shape for each beam stopping position. Similarly, beam weights were optimized at each beam position within individual planes to achieve intensity modulation.
Fig. 1. Each individual beam is designed to treat only the outer-most pleural surface as it is projected from the incident angle (A). As more fields are added (B), a larger portion of the pleura is included within the treatment volume.
RESULTS Review of the combination photon/electron technique demonstrates a number of critical faults. As described, two-thirds of the lung volume receives at least 50% of the dose and almost half receives more than 60%. Thus, the total dose prescribed to the pleural surface is necessarily limited to 4250 cGy. This dose is likely inadequate for a reasonable probability of local control or cure. The isodose distributions represented also show that adequate coverage of the medial portion on the pleural lining is virtually impossible to achieve (treated to below 50% dose) if any significant lung sparing is to be maintained. These underdosed pleural surfaces represent areas that could ultimately result in treatment failure. A representative plane (Fig. 3) shows the potential for dose coverage achievable by using the IMRT technique. This plane shows that the 100% isodose value has been precisely shaped to create a “shell” of dose that treats only the desired pleural surface with included margin. Centrally, lung parenchyma sparing has been achieved, with two-thirds of the lung receiving less than 50% of the dose and approximately half of the lung receiving less than 25%. Heart and spinal cord doses were not specifically discussed for the photon/electron technique; however,
Intensity modulated treatment of pleural mesothelioma ● M. TOBLER et al.
257
Dose distribution to normal tissues outside the lung was analyzed for the IMRT technique. Optimization of beam shapes, trajectories, and relative weighting allows for significant sparing of the spinal cord, with a maximum dose of 50% being delivered. The heart is also relatively spared with the IMRT technique, receiving no greater than 25% of the prescribed dose to a small volume. This compares very favorably with the mixed photon/electron technique. While not represented, one drawback to the IMRT technique is that the positional changes in beam direction result in dose delivered to the opposite lung. The maximum dose is approximately 25%, delivered to small strips of the contralateral lung. Figure 4 shows the relative field weightings applied for creation of the IMRT plan shown in Fig. 3. Unconstrained least-squares optimization was applied to create the isodose distribution represented. DISCUSSION Fig. 2. The simulated block for 1 fixed-gantry position traverses the entire superior and inferior extent of the projected outer pleural edge.
isodose evaluation shows that, while totally uninvolved, the spinal cord is included in the full-dose region of the photon fields and at least half of the heart is also included.
The most widely accepted technique for treating pleural mesothelioma with an intact lung is the mixed photon/electron technique described by Kutcher et al.1 As previously mentioned, the dose inhomogeneity seen as a result of the required overlap between the photon and electron fields is undesirable, yet unavoidable. This technique also has a more difficult time covering the entire pleural surface while still maintaining adequate lung sparing due to the sloping surface of the pleura that
Fig. 3. A representative isodose distribution for a single plane shows the dramatic improvement in delivered dose and tissue sparing achievable with the IMRT technique.
258
Medical Dosimetry
Volume 27, Number 4, 2002
4. a tool that would automatically increment the simulated gantry rotation at a specified degree setting.
Fig. 4. This is a graphical representation, for a single plane, of the intensities required from each individual beam (bottom) to create a more uniform delivery of dose (top).
must be covered within both the electron and photon fields. Because the electron portion of this technique is often unable to reach the deeper pleural tissues in the superior planes, these tissues must be treated by the anterior and posterior photon fields, resulting in delivery of the entire prescribed dose to this portion of the lung. Extra lung sparing in the superior direction can be gained in the IMRT technique because the same “shell”-type distribution can extend to the superior pleural lining by continued optimization of blocking and beam weighting through those planes. This results in an additional sparing of lung in the superior direction that would be unachievable with photon/electron technique. During creation of the blocking for this technique, several additional beam entry and modification tools were identified. These include: 1. a visualization tool that would allow simultaneous display of the beam’s-eye-view projection and sequential axial views; 2. a placement tool allowing beam definition in either the beam’s-eye or axial view; 3. an automation process for leaf position following manual placement of the beam edge, as well as a way to automatically set the margin of the beam around the desired outer lung volume as it is projected at the various rotational angles; and
We are currently developing an optimization routine to assist in the adjustment block positions and beam weights (i.e., individual leaf positions and required dynamic adjustments needed per number of monitor units delivered). Constrained optimization will need to be developed that will automatically optimize the beams for each individual leaf position, for each individual beam, in each individual plane, while simultaneously minimizing dose to the treated lung, the opposite lung, and to the adjacent normal tissue. This technique works best in concave lung segments, as around the anterior, posterior, and lateral chest. However, in convex lung segments, such as adjacent to the mediastinum, the photon beams directed tangentially to the pleural surface must transit a greater volume of lung. This increases the lung dose compared to that in sections of the lung having completely concave geometry. Improvement in the dose distribution is seen as the number of simulated treatment fields increases, where the best distribution should result from 360 different fixed beams directed from 360 different positions (360° rotation). While the description of this technique is most similar to that of a rotational-field technique, there are some definite differences. Due to the difference in beam weightings along the length of the beam at any given angle, ultimate delivery would occur in an automated step-and-shoot fashion. Treatment would start at a given fixed position. As the monitor units are delivered, leaf positions, corresponding to the different field weightings determined for each individual plane, would automatically adjust until the entire dose from that field had been delivered. Once complete, the gantry would rotate to the next fixed position and begin the process for the next field. The treatment planning examples shown are idealized because they neglect the impact of normal respiration during treatment. This is a serious problem that must be addressed through creative techniques such as active breathing control pioneered by Wong et al.,3 Stromberg et al.,4 and Willett et al.,5 and possibly through use of margins of greater than 5 mm to reflect motion in shallow breathing.
CONCLUSIONS This rotational IMRT treatment technique shows the potential for dramatic improvement of dose delivered to the lung, spinal cord, and heart when compared to the currently available technique. While we have identified additional tools that will allow better dose optimization for this technique, the doses delivered are achievable through application of technologies that are currently available to the treatment planner.
Intensity modulated treatment of pleural mesothelioma ● M. TOBLER et al.
REFERENCES 1. Kutcher, J.G.; Kestler, C.; Greenblatt, C.; et al. Technique for external beam treatment of mesothelioma. Int. J. Radiat. Oncol. Biol. Phys. 13:1747–52; 1987. 2. Mattson, K.; Holsti, L.R.; Tammilehto, L.; et al. Multimodality treatment programs for malignant pleural mesothelioma using highdose hemithorax irradiation. Int. J. Radiat. Oncol. Biol. Phys. 24: 643–50; 1992. 3. Wong, J.W.; Sharpe, M.B.; Jaffray, D.A.; et al. The use of active
259
breathing control (ABC) to reduce margin for breathing motion. Int. J. Radiat. Oncol. Biol. Phys. 44:911–99; 1999. 4. Stromberg, J.S.; Sharpe, M.D.; Kim, L.H.; et al. Active breathing control (ABC) for Hodgkin’s disease: Reduction in normal tissue irradiation with deep inspiration and implications for treatment. Int. J. Radiat. Oncol. Biol. Phys. 48:797–806; 2000. 5. Willett, C.G.; Linggood, R.M.; Stracher, M.A.; et al. The effect of the respiratory cycle on mediastinal and lung dimensions in Hodgkin’s disease. Implications for radiotherapy gated to respiration. Cancer 60:1232–7; 1987.
PII: S0958-3947(02)00149-8
INTENSITY-MODULATED PHOTON ARC THERAPY FOR TREATMENT OF PLEURAL MESOTHELIOMA MATT TOBLER, C.M.D., GORDON WATSON, M.D., PH.D., and DENNIS D. LEAVITT, PH.D. University of Utah Health Science Center, Salt Lake City, UT (Accepted 15 April 2001)
Abstract—Radiotherapy plays a key role in the definitive or adjuvant management of patients with mesothelioma of the pleural surface. Many patients are referred for radiation with intact lung following biopsy or subtotal pleurectomy. Delivery of efficacious doses of radiation to the pleural lining while avoiding lung parenchyma toxicity has been a difficult technical challenge. Using opposed photon fields produce doses in lung that result in moderate-to-severe pulmonary toxicity in 100% of patients treated. Combined photon-electron beam treatment, at total doses of 4250 cGy to the pleural surface, results in two-thirds of the lung volume receiving over 2100 cGy. We have developed a technique using intensity-modulated photon arc therapy (IMRT) that significantly improves the dose distribution to the pleural surface with concomitant decrease in dose to lung parenchyma compared to traditional techniques. IMRT treatment of the pleural lining consists of segments of photon arcs that can be intensity modulated with varying beam weights and multileaf positions to produce a more uniform distribution to the pleural surface, while at the same time reducing the overall dose to the lung itself. Computed tomography (CT) simulation is critical for precise identification of target volumes as well as critical normal structures (lung and heart). Rotational arc trajectories and individual leaf positions and weightings are then defined for each CT plane within the patient. This paper will describe the proposed rotational IMRT technique and, using simulated isodose distributions, show the improved potential for sparing of dose to the critical structures of the lung, heart, and spinal cord. © 2002 American Association of Medical Dosimetrists. Key Words:
Treatment planning, Intensity modulation, Arc, Mesothelioma.
the lung, deliver doses of 4250 to the 100% isodose level but block out the central lung volume. Anterior and posterior electron fields deliver the same dose to the untreated pleural lining, which is blocked from the anterior and posterior photon fields and is treated to the 90% isodose value. One difficulty encountered when treating with this technique is that it often requires an overlap between the photon and electron fields, creating an undesired high-dose region in the superficial chest wall at the peripheral edges of the electron fields. Secondarily, selection of the appropriate electron energy becomes critical. Electrons have a more difficult time covering the lining of the lung laterally as it begins to slope away from the beam, suggesting the need for use of a higher energy electron beam. In the mid lung region, the lung rises closer to the chest surface, suggesting that a slightly lower energy electron beam be used. It is possible to resolve this problem, however, by choosing the higher of the electron energies to produce the desired lateral coverage and then add compensation to reduce doses delivered to the mid portion of the lung. The falloff dose from these higher energy electrons, however, again results in added undesired dose delivery to the lung. Consequently, patients have a high incidence of pneumonitis as a major complication of therapy. In this work, we propose the use of a new technique that will allow delivery of a more homogeneous dose to the entire pleural lining while at the same time producing better sparing of the healthy lung volume, as well as the
INTRODUCTION Mesothelioma is a tumor that involves the entire outer surface of the lung (pleural lining). The goal of treating patients with this disease is to deliver curative doses (usually ⱖ 4250 cGy) to the entire pleural surface on the involved side, while at the same time minimizing the dose delivered to the uninvolved tissues of the lung itself and surrounding uninvolved structures. Previous published techniques1,2 include both treatment of the entire involved lung with anterior and posterior photon beams and a combination photon/electron beam approach. Treatment with opposed photon beams that treat the entire lung volume results in 1 of 2 outcomes: (a) excessive dose is delivered to the entire lung, resulting in loss of function; or (b) decrease in the delivery of dose to the pleural lining, resulting in inadequate treatment of the desired volume. Treatment using similar fields and a combination of photons and electrons allows the treatment planner to decrease the dose delivered to the lung by placing a carefully selected electron beam directly incident the lung volume. This should allow better limitation of doses delivered to the healthy uninvolved portion of the lung. In this case, the photon fields, which treat the medial, lateral, superior, and inferior aspects of Reprint requests to: M. Tobler, University of Utah, Department of Radiation Oncology, Health Science Center, 50 North Medical Drive, Salt Lake City, UT 84132. E-mail: matthew.tobler@ hsc.utah.edu 255
256
Medical Dosimetry
Volume 27, Number 4, 2002
uninvolved normal structures, such as the heart and spinal cord. This technique combines the use of dynamic multileaf collimations and dose delivery with multiple narrow-beam rotational field projections to producing a “shell” type isodose distribution that treats the entire pleural surface while maintaining a “cold” central core within the lung volume itself. Isodose distributions will be used to illustrate the potential gains in dose uniformity with this technique compared to the combination photon/ electron technique. METHODS AND MATERIALS When treating with the previously described photon/electron technique, the patient may be positioned with his/her arms by his/her sides because the anterior and posterior photon treatment fields can be easily shaped to exclude treatment of the arms. Due to the rotational nature of our dynamic IMRT technique, however, patient data was obtained and evaluated with the patient’s arms in an upright position. For this comparative patient data set, CT scans were obtained with 0.5-cm spacing extending from 3 cm above the superior extent of the lung to 3 cm below the diaphragm. The CT scans were sent to the RAHD treatment planning computer (RAHD Oncology Products, St. Louis, MO). Structures of interest were outlined, including both lungs, the heart, and the spinal cord. For comparison, we will refer to the photon/electron technique previously described in the literature.1 CT information and treatment planning was used to evaluate (a) the desired electron energies, (b) the potential need for placement of compensation in the electron fields, and (c) the possible need for overlap between the photon and electron fields. The anterior and posterior photon fields were designed to block out the central portion of the lung to a margin of 1 cm inside the projected lung volume in all directions. Blocks were also designed to exclude portions of the liver and stomach. Electron fields were matched to the blocked portion of the photon fields in the lung, and included a slight overlap to ensure adequate coverage of the desired volume. As previously mentioned, with this technique, dose to the pleural surface was limited to the 90% value for the electron fields. For creation of the IMRT technique, the isocenter was placed in the central portion of the lung to allow for rotation of the beam around the entire lung volume. Beams were created at fixed positions every 10° of gantry rotation (Fig. 1). Blocking for each individual beam trajectory was created so that each plane had a 0.5-cm margin both inside and outside the pleural lining of the lung to produce the desired volume coverage (Fig. 2). Each beam projection was independently evaluated through beam’s-eye view to allow for adjustment of the field shape for each beam stopping position. Similarly, beam weights were optimized at each beam position within individual planes to achieve intensity modulation.
Fig. 1. Each individual beam is designed to treat only the outer-most pleural surface as it is projected from the incident angle (A). As more fields are added (B), a larger portion of the pleura is included within the treatment volume.
RESULTS Review of the combination photon/electron technique demonstrates a number of critical faults. As described, two-thirds of the lung volume receives at least 50% of the dose and almost half receives more than 60%. Thus, the total dose prescribed to the pleural surface is necessarily limited to 4250 cGy. This dose is likely inadequate for a reasonable probability of local control or cure. The isodose distributions represented also show that adequate coverage of the medial portion on the pleural lining is virtually impossible to achieve (treated to below 50% dose) if any significant lung sparing is to be maintained. These underdosed pleural surfaces represent areas that could ultimately result in treatment failure. A representative plane (Fig. 3) shows the potential for dose coverage achievable by using the IMRT technique. This plane shows that the 100% isodose value has been precisely shaped to create a “shell” of dose that treats only the desired pleural surface with included margin. Centrally, lung parenchyma sparing has been achieved, with two-thirds of the lung receiving less than 50% of the dose and approximately half of the lung receiving less than 25%. Heart and spinal cord doses were not specifically discussed for the photon/electron technique; however,
Intensity modulated treatment of pleural mesothelioma ● M. TOBLER et al.
257
Dose distribution to normal tissues outside the lung was analyzed for the IMRT technique. Optimization of beam shapes, trajectories, and relative weighting allows for significant sparing of the spinal cord, with a maximum dose of 50% being delivered. The heart is also relatively spared with the IMRT technique, receiving no greater than 25% of the prescribed dose to a small volume. This compares very favorably with the mixed photon/electron technique. While not represented, one drawback to the IMRT technique is that the positional changes in beam direction result in dose delivered to the opposite lung. The maximum dose is approximately 25%, delivered to small strips of the contralateral lung. Figure 4 shows the relative field weightings applied for creation of the IMRT plan shown in Fig. 3. Unconstrained least-squares optimization was applied to create the isodose distribution represented. DISCUSSION Fig. 2. The simulated block for 1 fixed-gantry position traverses the entire superior and inferior extent of the projected outer pleural edge.
isodose evaluation shows that, while totally uninvolved, the spinal cord is included in the full-dose region of the photon fields and at least half of the heart is also included.
The most widely accepted technique for treating pleural mesothelioma with an intact lung is the mixed photon/electron technique described by Kutcher et al.1 As previously mentioned, the dose inhomogeneity seen as a result of the required overlap between the photon and electron fields is undesirable, yet unavoidable. This technique also has a more difficult time covering the entire pleural surface while still maintaining adequate lung sparing due to the sloping surface of the pleura that
Fig. 3. A representative isodose distribution for a single plane shows the dramatic improvement in delivered dose and tissue sparing achievable with the IMRT technique.
258
Medical Dosimetry
Volume 27, Number 4, 2002
4. a tool that would automatically increment the simulated gantry rotation at a specified degree setting.
Fig. 4. This is a graphical representation, for a single plane, of the intensities required from each individual beam (bottom) to create a more uniform delivery of dose (top).
must be covered within both the electron and photon fields. Because the electron portion of this technique is often unable to reach the deeper pleural tissues in the superior planes, these tissues must be treated by the anterior and posterior photon fields, resulting in delivery of the entire prescribed dose to this portion of the lung. Extra lung sparing in the superior direction can be gained in the IMRT technique because the same “shell”-type distribution can extend to the superior pleural lining by continued optimization of blocking and beam weighting through those planes. This results in an additional sparing of lung in the superior direction that would be unachievable with photon/electron technique. During creation of the blocking for this technique, several additional beam entry and modification tools were identified. These include: 1. a visualization tool that would allow simultaneous display of the beam’s-eye-view projection and sequential axial views; 2. a placement tool allowing beam definition in either the beam’s-eye or axial view; 3. an automation process for leaf position following manual placement of the beam edge, as well as a way to automatically set the margin of the beam around the desired outer lung volume as it is projected at the various rotational angles; and
We are currently developing an optimization routine to assist in the adjustment block positions and beam weights (i.e., individual leaf positions and required dynamic adjustments needed per number of monitor units delivered). Constrained optimization will need to be developed that will automatically optimize the beams for each individual leaf position, for each individual beam, in each individual plane, while simultaneously minimizing dose to the treated lung, the opposite lung, and to the adjacent normal tissue. This technique works best in concave lung segments, as around the anterior, posterior, and lateral chest. However, in convex lung segments, such as adjacent to the mediastinum, the photon beams directed tangentially to the pleural surface must transit a greater volume of lung. This increases the lung dose compared to that in sections of the lung having completely concave geometry. Improvement in the dose distribution is seen as the number of simulated treatment fields increases, where the best distribution should result from 360 different fixed beams directed from 360 different positions (360° rotation). While the description of this technique is most similar to that of a rotational-field technique, there are some definite differences. Due to the difference in beam weightings along the length of the beam at any given angle, ultimate delivery would occur in an automated step-and-shoot fashion. Treatment would start at a given fixed position. As the monitor units are delivered, leaf positions, corresponding to the different field weightings determined for each individual plane, would automatically adjust until the entire dose from that field had been delivered. Once complete, the gantry would rotate to the next fixed position and begin the process for the next field. The treatment planning examples shown are idealized because they neglect the impact of normal respiration during treatment. This is a serious problem that must be addressed through creative techniques such as active breathing control pioneered by Wong et al.,3 Stromberg et al.,4 and Willett et al.,5 and possibly through use of margins of greater than 5 mm to reflect motion in shallow breathing.
CONCLUSIONS This rotational IMRT treatment technique shows the potential for dramatic improvement of dose delivered to the lung, spinal cord, and heart when compared to the currently available technique. While we have identified additional tools that will allow better dose optimization for this technique, the doses delivered are achievable through application of technologies that are currently available to the treatment planner.
Intensity modulated treatment of pleural mesothelioma ● M. TOBLER et al.
REFERENCES 1. Kutcher, J.G.; Kestler, C.; Greenblatt, C.; et al. Technique for external beam treatment of mesothelioma. Int. J. Radiat. Oncol. Biol. Phys. 13:1747–52; 1987. 2. Mattson, K.; Holsti, L.R.; Tammilehto, L.; et al. Multimodality treatment programs for malignant pleural mesothelioma using highdose hemithorax irradiation. Int. J. Radiat. Oncol. Biol. Phys. 24: 643–50; 1992. 3. Wong, J.W.; Sharpe, M.B.; Jaffray, D.A.; et al. The use of active
259
breathing control (ABC) to reduce margin for breathing motion. Int. J. Radiat. Oncol. Biol. Phys. 44:911–99; 1999. 4. Stromberg, J.S.; Sharpe, M.D.; Kim, L.H.; et al. Active breathing control (ABC) for Hodgkin’s disease: Reduction in normal tissue irradiation with deep inspiration and implications for treatment. Int. J. Radiat. Oncol. Biol. Phys. 48:797–806; 2000. 5. Willett, C.G.; Linggood, R.M.; Stracher, M.A.; et al. The effect of the respiratory cycle on mediastinal and lung dimensions in Hodgkin’s disease. Implications for radiotherapy gated to respiration. Cancer 60:1232–7; 1987.