- Email: [email protected]
334 poster Management with three-dimensional CT-guided pararectal brachytherapy of patients with local recurrence who failed initial brachytherapy
Posters
St 15
degrees. The collimator jaws used conventionally are redundant and therefore removed. Blocking equivalent to a conventional MLC and back-up collimator set-up is achieved by the combination of three layers of leaves. For the delivery of an IMRT fluence a new sequencer was developed. The sequencer for each layer of banks works recursively. A segment is derived from the input fluence. It's contribution is subtracted and the remaining fluonce is used as input for the next segment. For this process in principle the rules for conventional sequencing can be applied (close-in, sliding window etc.). However, if the generation of a segment is an independent process for the three layers of banks, it is possible that the three layers produce nonoverlapping shapes. Therefore, a connection needs to be established between the sequencing processes of the three layers to ensure that all three banks work together in creating one open segment. The connection between the sequencing processes is made via masks. The segment created by the first layer of banks will serve as input to the sequencer for the second layer of banks. This mask in combination with the fluence distribution is used to generate the segment shape for the second layer. Similarly, the segment created by the first and second layers will be used as mask for the third layer. A concave shape deliverable by one of the layers may not be deliverable by another. In such a case the segment witl be split into to segments. Fine-tuning of the leaf position of the other banks used to improve the boundary definition of a segment and optimal blocking of the area outside the segment by as many as possible layers of leaves. The performance of the sequencer is demonstrated in clinical examples. We acknowledge Elekta for their support. 331
poster
Reduction in lung tumour motion using active breathing control J.A. Christian 1, H.A. McNair2, J.R.N. Symonds-Tay/er 1, C. Knowles2, J.L. Bedford2, M. Brada 1 1institute of Cancer Research, Academic Department of Radiotherapy, Sutton, United Kingdom 2Royal Marsden Hosptial, Department of Radiotherapy, Sutton, United Kingdom Backaround: With the advent of high precision radiotherapy for lung cancer, the issue of tumour movement has become increasingly important. Lung tumour movement is related not only to diaphragmatic movement but also cardiac and aortic pulsations. We have piloted the use of an Active Breathing Control (ABC) device, (William Beaumont Hospital, Detroit, USA) to study the feasibility of use and the effects on lung tumour movement in a series of patients undergoing radical radiotherapy for non-small cell lung cancer (NSCLC). Methods: Patients were set up in the treatment position in the simulator, their tumour identified and movement measured in 3-D using simulator wires and fluoroscopy to mark extreme positions of movement from a baseline position. Patients were trained to use the ABC device, gate levels set and repeat measurements of baseline and maximal tumour movement were made. Tumour movements with and without the ABC device were compared using a Student's t-test. The reproducibility of tumour position with each breathhold was measured using couch correction. Results: Ten patients with mean age 69.2 years (range 46 - 82 years) were assessed; all were performance status 0 or 1 and were undergoing radical radiotherapy for NSCLC. Five patients were medically unfit for surgery and two had a forced expiratory volume in one second of < 1.0 I. Tumour stage ranged from Stage 1A to 3B. Mean reduction in tumour movement with the ABC device was superior/inferior 9.0 mm (range 1 - 21 mm) p = 0.004; left/right 3.7 mm (range 0 - 6 mm) p = 0.003; anterior/posterior 4.1 mm (range 0 - 8 mm) p = 0.003. Mean residual tumour movement in a superior/inferior direction was 0.3 mm (range 0 - 1 mm); left/right 0.8 mm (range 0 - 3 mm); anterior/posterior 0.7 mm (range 0 - 2 mm). Residual movement was cardiac related. Reproducibility studies were performed for nine patients; tumour position with the ABC device was reproducible with each breathhold in seven. The two other patients required couch corrections of 1 mm and 2 mm respectively. Conclusion: ABC is an effective and well-tolerated method for reducing tumour movement in lung cancer patients. We intend to introduce the ABC device into conformal radiotherapy treatments for lung cancer to improve accuracy of delivery and reduce planning target volume margins.
332
poster
MRI-based 3DCRT and IMRT H. Shukla, P. Klahr, P. Seltzer, D. Abraham Philips Medical Systems, PROS I CT, Cleveland, USA Purpose: External beam 3D conformal RTP based solely on MRI volumes has a number of advantages over CT, or even Fusion with respect to efficiency.[Shukla, et.al., AAPM 1999] However, sole use of MRI requires a high degree of confidence in the geometric accuracy of the images. Once qualified as acceptable for 3D-CRT, we examine the possibility of extending useage to IMRT. Methods and Materials: Sequential pelvic CT (PQ5000, Philips) and distortion-corrected MRI (Panaraoma 023T, Philips) scans were obtained with similar volumes, resolution, and patient immobilization. 3D-CRT and IMRT plans were generated (Pinnacle, Philips) for both CT and MRI, using consistent segmentation definitions for Fusion. Results: For 3D-CRT plans, results with CT and MRI were similar. IMRT plans were more problematic for evaluation, but also appeared comparable. Conclusion: External beam planning using MRI solely appears to be a promising tool, and may possible be extended to IMRT. Improved metrics for comparison are desired for future evaluation, as well as the impact of homogeneity corrections (vs. bulk density) in the pelvis, abdomen, and brain. 333
poster
Computer aided treatment planning for total body irradiation under translation conditions J. Licher 1, I. Mansi2, J. M(~ller2, U. Ramm 1, S. Sch~fer 1, F. Rudolf 1, J. Moog 1, C. Rahl 1, H.D. BSttcher 1, S. Mose 1 1Klinikum der Johann Wolfgang Goethe-Universitbt, Klinik for Strahlentherapie und Onkotogie, Frankfurt am Main, Germany 2Fachhochschule GieBen-Friedberg, Fachbereich Krankenhaus- und Medizintechnik, Umwelt- und Biotechnologie, GieBen, Germany Purpose: The evaluation of treatment parameters (speed of translation, thickness of the lung protection) for the total body irradiation (TBI) under translation conditions is based on the calculation of the dose distribution. Conventionally, the parameters are acquired manually by measuring geometrical patient parameters using cost consuming sheets of CT-images. Presently, there is no commercial treatment planning system available to calculate the treatment parameters. A prototype of a new computer aided planning system is presented, providing a film free method for the determination of the treatment parameters for TBI under translation conditions. Method: The new TBI planning system was developed using the Matlab@Environment. The required patients CT-Images are transferred by a DICOM-interface to the new planning workstation. Then the CT-Images are analyzed by this program by a semi automatic operation and the treatment parameters are computed. Additionally, doses at relevant points of the body are calculated. The verfication of the new TBI plannning system was performed using Alderson phantoms (male and female). Comparisons of the treatment parameters calculated by the new TBI planning system and the conventional technique were done. Results: The computer aided determination of the treatment parameters is able to reduce the time and costs considerably. There is no need for exposing sheets of CT-Images for treatment planning process. The comparisons based on CT-Images of Alderson phantoms have shown the excellent accuracy of this new system. For determination of geometrical parameters both techniques agree within few millimeters. The differences of the computed and conventionally calculated treatment parameters are negligible. ~onclusion: The treatment planning system prototype based on the Matlab@-Environment provides an accurate, convenient, efficient and cost saving way for the calculation of the necessary treatment parameters for the total body irradiation under translation conditions. The evalutation for using the planning system in clinical routing are still in progress. 334
poster
Management with three-dimensional CT-guided pararectal brachytherapy of patients with local recurrence who failed initial brachytherapy P. Koutrouvelis, F. Hendricks, N. Lailas, G. GiI-Montero, J. Sehn, N. Khawand, H. Bondy, S. Katz Uro-Radiology Prostate Institute, Brachytherapy, Vienna, Virginia, U.S.A. Purpose: 3D CT-guided diagnosis and treatment of local recurrent adenocarcinoma of the prostate after failed brachytherapy.
$t 16
Posters
Methods and Materials: Thirty-one (31) patients had biochemical failure one to four years after brachytherapy. Patients age range was 51 - 79 years. Twelve patients had Gleason's score < 6, twelve = 7 and seven > 8. Initial PSA range was 4.6 to 63 ng/mL. Nine patients had clinical stage T1 - T2a and twenty-two were T2b or T3a. 1251 seeds were used in 5 patients (16%) and 103Pd in 26 patients (84%). All patients underwent 3-D CT-guided pararectal extensive biopsy under local anesthesia or IV sedation. Eleven of the thirty-one had invasion of one or both seminal vesicles that had not been diagnosed or treated with initial brachytherapy and twenty had residual/recurrent prostate cancer. All patients had normal CT of the upper and lower abdomen and normal nuclear bone scan. Only the area of positive biopsy was treated with 3-D CT-guided pararectal brachytherapy. 1251 seeds in strand were used in 24 patients (77%) and 103Pd in 7 patients (23%). The activity of 1251 seeds was in the range of 0.35 to 0.38 mCi/seed and dosage was 100 - 144 Gy to the recurrence site. The activity of 103Pd was in the range of 1.20 - 1.40 mCi/seed, and dosage was 100 - 120 Gy to the recurrence site. All patients except one had three months of neoadjuvant androgen ablation. Follow-up was 1 - 6 years (median, 3 years) in 30 patients. Result: A high level of biochemical control (90%) was achieved in all of these patients who are recognized as high risk due to local recurrence. Four patients experienced grade 2 or 3 GI or GU complications and two patients experienced grade 4 GI complications. Conclusion: Computed tomography imaging and three-dimensional computed tomography directed needle biopsy is an effective means to identify local recurrence in patients with biochemical failure after brachytherapy. Those patients with local recurrent disease including seminal vesicle invasion can then be successfully treated with pararectal three-dimensional CTguided supplemental brachytherapy to the site of the recurrence. Assessment of the seminal vesicle status is an essential part of staging for local recurrence. 335
poster
Optimisation of radiotherapy and multimodality therapy for cervical cancer patients with poor prognosis I. Kasenka N.N. Alexandrov Research Institute of Oncology and Medical Radiology, Oncology/Gynaecology, P.O. Lesnoy-2, Minsk Belarus Purpose of the study: to enhance the efficacy of the treatment of cervical cancer patients with poor prognosis developing new techniques of antitumor therapy. Object of the study: 557 patients with locally advanced cervical cancer. Summary of results: individualisation of methodological approaches to administration of split-course combination radiotherapy increases the therapeutic interval. At the same time the number of complications decreases 2-fold, 5-year survival rate of patients with T3 tumours is significantly higher (by 32.7%), 20-year survival- by 31%, local recurrence rate is reduced 3-fold in comparison with the conventional course of combination radiotherapy. The advantage of 137-Cs radionuciide in isoeffective fractionation regimens is in the reduction of radiation morbidity and complication rates 3 and 2 times respectively, compard to 60-Co. Supplementary epithalamin (100 mg) or thymalin (100 mg biotherapy in combination radiotherapy schemes makes it possible to decrease the rate of radiation morbidity 2.3 - 3-fold, complications - 1.5 - 2-fold. The multimodal treatment including intra-arterial multidrug chemotherapy (5-fluorouracil+cisplatin), epithalamin biotherapy, split-course radiotherapy and ajuvant systemic multidrug chemotherapy significantly improves 3-year survival rate (by 20.8%) of T3 cervical cancer patients, compared to radiotherapy alone, the rate of distant metastasising in reduced 2-fold. 336
poster
The speed modulation of a patient's rotation in the rotarydual technique of the total skin electron irradiation. In-phantom results 7". Piotrowski Greatpoland Cancer Center, Department of Medical Physics, Poznan, Poland Purpose: The aim of this study was to reduce the differences in doses accumulated in the points, Iocalised in the same horizontal plane, on the patient's skin, during the rotary-dual total skin electron irradiation (TSEI RD). Material and Method: During the standard TSEI RD technique, the patient was standing on the slowly and continuously rotating platform while the irradiation with two dual electron fields was performed. In this study a non-continuous platform movement was implemented. One cylindrical (diameter of
30 cm) and two elliptical (diameters of 20/40 cm and 25/35 cm) wax phantoms were used. The one full circle (360 degree) of the platform was divided into eight fractions. The platform speed was calculated individually for the each fraction. The total time of one rotation in both modes of the movements: continuous and non-continuous were equal. Phantoms were irradiated with two dual fields: size of 36x36 cm at the isocenter each, sourcephantom distance of 300 cm, and with 6 MeV electrons' energy (output). Doses were measured with the semiconductor detectors which were placed equidistantly (distance between dosimetrical points by 45 degree) on the same horizontal plane on the phantom's surface. Results: The respective measured doses after normalisation to those previously calculated were as follows: 1/for the cylindrical phantom (diameter of 30 cm) - 100.1%, 99.8%, 100.3%, 99.4%, 100.1%, 99.8%, 100,3%, 99.4%; 2/for the elliptical phantom of 25/35 cm - 98.3%, 99.6%, 103.2%, 99.2%, 97.8%, 99.6%, 103.2%, 100.7% and 3/for the elliptical phantom of 20/40 cm - 95.8%, 99.9%, 105%, 98.8%, 95.8%, 100.2%, 105%, 99.3%. The modulation of the speed during platform's rotation resulted in the decrease of the dose discrepancies. The respective measured doses after the normalization performed with the same as above way were equal to: 1/ for the elliptical phantom of 25/35 cm - 99.4%, 99.9%, 101.3%, 100.5%, 100.1%, 100.9%, 101 1%, 99.7%; 2/for the elliptical phantom of 20/40cm 98.9%, 100.2%, 102.4%, 99.8%, 99%, 99.9%, 102.1%, 100.5%. Conclusion: Results of the dose in-phantom measurements showed that the speed modulation of the platform reduced the dose inhomogeneity at the horizontal plane on the phantom's surface.
STEREOTACTIC 337
poster
Dosimetric analysis of the dynamic conformal arc technique for stereotactic radiotherapy of pituitary tumours K.L. James 1, J.L. Robat2, B.G. Clark2 1BC Cancer Agency, Radiation Therapy, Vancouver, Canada 2BC Cancer Agency, Medical Physics, Vancouver, Canada PurDose: With the advent of micro-multileaf collimation, improved target conformity in stereotactic radiotherapy (SRT) can be achieved using an arrangement of static conformal beams. While this approach usually yields acceptable target conformity, in comparison to traditional stereotactic arcs a higher peripheral dose occurs along the beam entrance paths. This work considers the dosimetric advantage of dynamic conformal arcs, whereby the beam aperture adapts to the target projection continuously through the arc. Treatment of pituitary tumours was chosen as a clinically-relevant but demanding test case because i) the target volume is centrally located; ii) immediately-proximal structures such as optic chJasm and brainstem must be spared; and iii) the dose to more peripheral structures such as temporal lobes must be minimized. Methods: Twenty-five SRT patients with pituitary tumours were re-planned using a) a symmetric 7-field static conformal beam arrangement, b) 5 equally-weighted dynamic conformal arcs, and c) 5 dynamic conformal arcs with increased weighting from the superior arc. Target coverage was specified as 99% of the target volume receiving 90% of the prescribed isocentre dose of 50 Gy. Treatment plans were evaluated quantitatively by comparing conformity index and dose homogeneity within the planning target volume. The advantage of dynamic arcs was assessed by determining the dose range over which sparing of normal tissue and structures was improved. Results: For approximately 80% of treatment plans, dynamic arcs yielded improved conformity to the target volume. Dose homogeneity was similar for all plans. For the dose range between 3 and 11 Gy, dynamic arcs consistently treated a lower volume of normal brain, above which sparing was comparable. Dynamic arcs achieved improved sparing of temporal lobes only in one-third of cases studied, for which larger target volumes resulted in significant overlap of static beams. By increasing the weight of the vertex dynamic arc, sparing of temporal lobes was improved for all plans at the expense of an increased volume of brainstem treated. For 70% of the plans, equally-weighted dynamic arcs produced improved sparing of brainstem at higher dose levels compared to static beams. Dose to the optic chiasm was similar for all three planning approaches. Conclusions: Dynamic conformal arc treatment allows for excellent target conformity with improved distribution of dose among peripheral normal brain tissues.
St 15
degrees. The collimator jaws used conventionally are redundant and therefore removed. Blocking equivalent to a conventional MLC and back-up collimator set-up is achieved by the combination of three layers of leaves. For the delivery of an IMRT fluence a new sequencer was developed. The sequencer for each layer of banks works recursively. A segment is derived from the input fluence. It's contribution is subtracted and the remaining fluonce is used as input for the next segment. For this process in principle the rules for conventional sequencing can be applied (close-in, sliding window etc.). However, if the generation of a segment is an independent process for the three layers of banks, it is possible that the three layers produce nonoverlapping shapes. Therefore, a connection needs to be established between the sequencing processes of the three layers to ensure that all three banks work together in creating one open segment. The connection between the sequencing processes is made via masks. The segment created by the first layer of banks will serve as input to the sequencer for the second layer of banks. This mask in combination with the fluence distribution is used to generate the segment shape for the second layer. Similarly, the segment created by the first and second layers will be used as mask for the third layer. A concave shape deliverable by one of the layers may not be deliverable by another. In such a case the segment witl be split into to segments. Fine-tuning of the leaf position of the other banks used to improve the boundary definition of a segment and optimal blocking of the area outside the segment by as many as possible layers of leaves. The performance of the sequencer is demonstrated in clinical examples. We acknowledge Elekta for their support. 331
poster
Reduction in lung tumour motion using active breathing control J.A. Christian 1, H.A. McNair2, J.R.N. Symonds-Tay/er 1, C. Knowles2, J.L. Bedford2, M. Brada 1 1institute of Cancer Research, Academic Department of Radiotherapy, Sutton, United Kingdom 2Royal Marsden Hosptial, Department of Radiotherapy, Sutton, United Kingdom Backaround: With the advent of high precision radiotherapy for lung cancer, the issue of tumour movement has become increasingly important. Lung tumour movement is related not only to diaphragmatic movement but also cardiac and aortic pulsations. We have piloted the use of an Active Breathing Control (ABC) device, (William Beaumont Hospital, Detroit, USA) to study the feasibility of use and the effects on lung tumour movement in a series of patients undergoing radical radiotherapy for non-small cell lung cancer (NSCLC). Methods: Patients were set up in the treatment position in the simulator, their tumour identified and movement measured in 3-D using simulator wires and fluoroscopy to mark extreme positions of movement from a baseline position. Patients were trained to use the ABC device, gate levels set and repeat measurements of baseline and maximal tumour movement were made. Tumour movements with and without the ABC device were compared using a Student's t-test. The reproducibility of tumour position with each breathhold was measured using couch correction. Results: Ten patients with mean age 69.2 years (range 46 - 82 years) were assessed; all were performance status 0 or 1 and were undergoing radical radiotherapy for NSCLC. Five patients were medically unfit for surgery and two had a forced expiratory volume in one second of < 1.0 I. Tumour stage ranged from Stage 1A to 3B. Mean reduction in tumour movement with the ABC device was superior/inferior 9.0 mm (range 1 - 21 mm) p = 0.004; left/right 3.7 mm (range 0 - 6 mm) p = 0.003; anterior/posterior 4.1 mm (range 0 - 8 mm) p = 0.003. Mean residual tumour movement in a superior/inferior direction was 0.3 mm (range 0 - 1 mm); left/right 0.8 mm (range 0 - 3 mm); anterior/posterior 0.7 mm (range 0 - 2 mm). Residual movement was cardiac related. Reproducibility studies were performed for nine patients; tumour position with the ABC device was reproducible with each breathhold in seven. The two other patients required couch corrections of 1 mm and 2 mm respectively. Conclusion: ABC is an effective and well-tolerated method for reducing tumour movement in lung cancer patients. We intend to introduce the ABC device into conformal radiotherapy treatments for lung cancer to improve accuracy of delivery and reduce planning target volume margins.
332
poster
MRI-based 3DCRT and IMRT H. Shukla, P. Klahr, P. Seltzer, D. Abraham Philips Medical Systems, PROS I CT, Cleveland, USA Purpose: External beam 3D conformal RTP based solely on MRI volumes has a number of advantages over CT, or even Fusion with respect to efficiency.[Shukla, et.al., AAPM 1999] However, sole use of MRI requires a high degree of confidence in the geometric accuracy of the images. Once qualified as acceptable for 3D-CRT, we examine the possibility of extending useage to IMRT. Methods and Materials: Sequential pelvic CT (PQ5000, Philips) and distortion-corrected MRI (Panaraoma 023T, Philips) scans were obtained with similar volumes, resolution, and patient immobilization. 3D-CRT and IMRT plans were generated (Pinnacle, Philips) for both CT and MRI, using consistent segmentation definitions for Fusion. Results: For 3D-CRT plans, results with CT and MRI were similar. IMRT plans were more problematic for evaluation, but also appeared comparable. Conclusion: External beam planning using MRI solely appears to be a promising tool, and may possible be extended to IMRT. Improved metrics for comparison are desired for future evaluation, as well as the impact of homogeneity corrections (vs. bulk density) in the pelvis, abdomen, and brain. 333
poster
Computer aided treatment planning for total body irradiation under translation conditions J. Licher 1, I. Mansi2, J. M(~ller2, U. Ramm 1, S. Sch~fer 1, F. Rudolf 1, J. Moog 1, C. Rahl 1, H.D. BSttcher 1, S. Mose 1 1Klinikum der Johann Wolfgang Goethe-Universitbt, Klinik for Strahlentherapie und Onkotogie, Frankfurt am Main, Germany 2Fachhochschule GieBen-Friedberg, Fachbereich Krankenhaus- und Medizintechnik, Umwelt- und Biotechnologie, GieBen, Germany Purpose: The evaluation of treatment parameters (speed of translation, thickness of the lung protection) for the total body irradiation (TBI) under translation conditions is based on the calculation of the dose distribution. Conventionally, the parameters are acquired manually by measuring geometrical patient parameters using cost consuming sheets of CT-images. Presently, there is no commercial treatment planning system available to calculate the treatment parameters. A prototype of a new computer aided planning system is presented, providing a film free method for the determination of the treatment parameters for TBI under translation conditions. Method: The new TBI planning system was developed using the Matlab@Environment. The required patients CT-Images are transferred by a DICOM-interface to the new planning workstation. Then the CT-Images are analyzed by this program by a semi automatic operation and the treatment parameters are computed. Additionally, doses at relevant points of the body are calculated. The verfication of the new TBI plannning system was performed using Alderson phantoms (male and female). Comparisons of the treatment parameters calculated by the new TBI planning system and the conventional technique were done. Results: The computer aided determination of the treatment parameters is able to reduce the time and costs considerably. There is no need for exposing sheets of CT-Images for treatment planning process. The comparisons based on CT-Images of Alderson phantoms have shown the excellent accuracy of this new system. For determination of geometrical parameters both techniques agree within few millimeters. The differences of the computed and conventionally calculated treatment parameters are negligible. ~onclusion: The treatment planning system prototype based on the Matlab@-Environment provides an accurate, convenient, efficient and cost saving way for the calculation of the necessary treatment parameters for the total body irradiation under translation conditions. The evalutation for using the planning system in clinical routing are still in progress. 334
poster
Management with three-dimensional CT-guided pararectal brachytherapy of patients with local recurrence who failed initial brachytherapy P. Koutrouvelis, F. Hendricks, N. Lailas, G. GiI-Montero, J. Sehn, N. Khawand, H. Bondy, S. Katz Uro-Radiology Prostate Institute, Brachytherapy, Vienna, Virginia, U.S.A. Purpose: 3D CT-guided diagnosis and treatment of local recurrent adenocarcinoma of the prostate after failed brachytherapy.
$t 16
Posters
Methods and Materials: Thirty-one (31) patients had biochemical failure one to four years after brachytherapy. Patients age range was 51 - 79 years. Twelve patients had Gleason's score < 6, twelve = 7 and seven > 8. Initial PSA range was 4.6 to 63 ng/mL. Nine patients had clinical stage T1 - T2a and twenty-two were T2b or T3a. 1251 seeds were used in 5 patients (16%) and 103Pd in 26 patients (84%). All patients underwent 3-D CT-guided pararectal extensive biopsy under local anesthesia or IV sedation. Eleven of the thirty-one had invasion of one or both seminal vesicles that had not been diagnosed or treated with initial brachytherapy and twenty had residual/recurrent prostate cancer. All patients had normal CT of the upper and lower abdomen and normal nuclear bone scan. Only the area of positive biopsy was treated with 3-D CT-guided pararectal brachytherapy. 1251 seeds in strand were used in 24 patients (77%) and 103Pd in 7 patients (23%). The activity of 1251 seeds was in the range of 0.35 to 0.38 mCi/seed and dosage was 100 - 144 Gy to the recurrence site. The activity of 103Pd was in the range of 1.20 - 1.40 mCi/seed, and dosage was 100 - 120 Gy to the recurrence site. All patients except one had three months of neoadjuvant androgen ablation. Follow-up was 1 - 6 years (median, 3 years) in 30 patients. Result: A high level of biochemical control (90%) was achieved in all of these patients who are recognized as high risk due to local recurrence. Four patients experienced grade 2 or 3 GI or GU complications and two patients experienced grade 4 GI complications. Conclusion: Computed tomography imaging and three-dimensional computed tomography directed needle biopsy is an effective means to identify local recurrence in patients with biochemical failure after brachytherapy. Those patients with local recurrent disease including seminal vesicle invasion can then be successfully treated with pararectal three-dimensional CTguided supplemental brachytherapy to the site of the recurrence. Assessment of the seminal vesicle status is an essential part of staging for local recurrence. 335
poster
Optimisation of radiotherapy and multimodality therapy for cervical cancer patients with poor prognosis I. Kasenka N.N. Alexandrov Research Institute of Oncology and Medical Radiology, Oncology/Gynaecology, P.O. Lesnoy-2, Minsk Belarus Purpose of the study: to enhance the efficacy of the treatment of cervical cancer patients with poor prognosis developing new techniques of antitumor therapy. Object of the study: 557 patients with locally advanced cervical cancer. Summary of results: individualisation of methodological approaches to administration of split-course combination radiotherapy increases the therapeutic interval. At the same time the number of complications decreases 2-fold, 5-year survival rate of patients with T3 tumours is significantly higher (by 32.7%), 20-year survival- by 31%, local recurrence rate is reduced 3-fold in comparison with the conventional course of combination radiotherapy. The advantage of 137-Cs radionuciide in isoeffective fractionation regimens is in the reduction of radiation morbidity and complication rates 3 and 2 times respectively, compard to 60-Co. Supplementary epithalamin (100 mg) or thymalin (100 mg biotherapy in combination radiotherapy schemes makes it possible to decrease the rate of radiation morbidity 2.3 - 3-fold, complications - 1.5 - 2-fold. The multimodal treatment including intra-arterial multidrug chemotherapy (5-fluorouracil+cisplatin), epithalamin biotherapy, split-course radiotherapy and ajuvant systemic multidrug chemotherapy significantly improves 3-year survival rate (by 20.8%) of T3 cervical cancer patients, compared to radiotherapy alone, the rate of distant metastasising in reduced 2-fold. 336
poster
The speed modulation of a patient's rotation in the rotarydual technique of the total skin electron irradiation. In-phantom results 7". Piotrowski Greatpoland Cancer Center, Department of Medical Physics, Poznan, Poland Purpose: The aim of this study was to reduce the differences in doses accumulated in the points, Iocalised in the same horizontal plane, on the patient's skin, during the rotary-dual total skin electron irradiation (TSEI RD). Material and Method: During the standard TSEI RD technique, the patient was standing on the slowly and continuously rotating platform while the irradiation with two dual electron fields was performed. In this study a non-continuous platform movement was implemented. One cylindrical (diameter of
30 cm) and two elliptical (diameters of 20/40 cm and 25/35 cm) wax phantoms were used. The one full circle (360 degree) of the platform was divided into eight fractions. The platform speed was calculated individually for the each fraction. The total time of one rotation in both modes of the movements: continuous and non-continuous were equal. Phantoms were irradiated with two dual fields: size of 36x36 cm at the isocenter each, sourcephantom distance of 300 cm, and with 6 MeV electrons' energy (output). Doses were measured with the semiconductor detectors which were placed equidistantly (distance between dosimetrical points by 45 degree) on the same horizontal plane on the phantom's surface. Results: The respective measured doses after normalisation to those previously calculated were as follows: 1/for the cylindrical phantom (diameter of 30 cm) - 100.1%, 99.8%, 100.3%, 99.4%, 100.1%, 99.8%, 100,3%, 99.4%; 2/for the elliptical phantom of 25/35 cm - 98.3%, 99.6%, 103.2%, 99.2%, 97.8%, 99.6%, 103.2%, 100.7% and 3/for the elliptical phantom of 20/40 cm - 95.8%, 99.9%, 105%, 98.8%, 95.8%, 100.2%, 105%, 99.3%. The modulation of the speed during platform's rotation resulted in the decrease of the dose discrepancies. The respective measured doses after the normalization performed with the same as above way were equal to: 1/ for the elliptical phantom of 25/35 cm - 99.4%, 99.9%, 101.3%, 100.5%, 100.1%, 100.9%, 101 1%, 99.7%; 2/for the elliptical phantom of 20/40cm 98.9%, 100.2%, 102.4%, 99.8%, 99%, 99.9%, 102.1%, 100.5%. Conclusion: Results of the dose in-phantom measurements showed that the speed modulation of the platform reduced the dose inhomogeneity at the horizontal plane on the phantom's surface.
STEREOTACTIC 337
poster
Dosimetric analysis of the dynamic conformal arc technique for stereotactic radiotherapy of pituitary tumours K.L. James 1, J.L. Robat2, B.G. Clark2 1BC Cancer Agency, Radiation Therapy, Vancouver, Canada 2BC Cancer Agency, Medical Physics, Vancouver, Canada PurDose: With the advent of micro-multileaf collimation, improved target conformity in stereotactic radiotherapy (SRT) can be achieved using an arrangement of static conformal beams. While this approach usually yields acceptable target conformity, in comparison to traditional stereotactic arcs a higher peripheral dose occurs along the beam entrance paths. This work considers the dosimetric advantage of dynamic conformal arcs, whereby the beam aperture adapts to the target projection continuously through the arc. Treatment of pituitary tumours was chosen as a clinically-relevant but demanding test case because i) the target volume is centrally located; ii) immediately-proximal structures such as optic chJasm and brainstem must be spared; and iii) the dose to more peripheral structures such as temporal lobes must be minimized. Methods: Twenty-five SRT patients with pituitary tumours were re-planned using a) a symmetric 7-field static conformal beam arrangement, b) 5 equally-weighted dynamic conformal arcs, and c) 5 dynamic conformal arcs with increased weighting from the superior arc. Target coverage was specified as 99% of the target volume receiving 90% of the prescribed isocentre dose of 50 Gy. Treatment plans were evaluated quantitatively by comparing conformity index and dose homogeneity within the planning target volume. The advantage of dynamic arcs was assessed by determining the dose range over which sparing of normal tissue and structures was improved. Results: For approximately 80% of treatment plans, dynamic arcs yielded improved conformity to the target volume. Dose homogeneity was similar for all plans. For the dose range between 3 and 11 Gy, dynamic arcs consistently treated a lower volume of normal brain, above which sparing was comparable. Dynamic arcs achieved improved sparing of temporal lobes only in one-third of cases studied, for which larger target volumes resulted in significant overlap of static beams. By increasing the weight of the vertex dynamic arc, sparing of temporal lobes was improved for all plans at the expense of an increased volume of brainstem treated. For 70% of the plans, equally-weighted dynamic arcs produced improved sparing of brainstem at higher dose levels compared to static beams. Dose to the optic chiasm was similar for all three planning approaches. Conclusions: Dynamic conformal arc treatment allows for excellent target conformity with improved distribution of dose among peripheral normal brain tissues.