- Email: [email protected]
High dose rate brachytherapy boost in the treatment of extremity soft-tissue sarcomas
Proceedings of the 46th Annual ASTRO Meeting
Conclusions: Brachytherapy provided a high level of local control for small tumours. Carcinologic, functional and cosmetic results were very good when implants were used as the exclusive treatment. Brachytherapy is a therapeutic alternative for peri-orificial face skin carcinomas because of the irregularity of contours and the difficulty in surgical reconstruction in this areas.
2403
A Phase I Hyperthermia- Induced Interleukin- 12 Gene- Therapy Trial in a Spontaneously Arising Feline Soft Tissue Sarcoma Model
F. Siddiqui,1 P. R. Avery,2 S. M. LaRue,1 M. L. Hauck,3 M. W. Dewhirst,4 C. Li,4 R. L. Ullrich1 Environmental and Radiological Health Sciences, Colorado State University, Fort Collins, CO, 2Department of Microbiology, Immunology and Pathology, Colorado State University, Fort Collins, CO, 3Department of Clinical Sciences, College of Veterinary Medicine, Raleigh, NC, 4Department of Radiation Oncology, Duke University, Durham, NC 1
Purpose/Objective: Cytokine genes are the most widely and extensively studied immunostimulatory agents in cancer gene therapy. In several studies, interleukin-12 (IL-12) was the most effective cytokine in inducing the eradication of experimental tumors, preventing development of metastases, and eliciting long-term antitumor immunity. Depending on the tumor model, IL-12 can exert anti- tumor activities via T cells, NK cells, or NKT cells. Induction of cytokines, such as IFN-g and IFN-inducible protein-l0, has also been implicated as a mechanism of antitumor activity of IL-12. Local and systemic administration of IL-12 protein has been studied in various murine models and in Phase I/II human trials. However, IL-12 protein therapy has been limited by dose-dependent toxicity. Local and efficient expression of IL-12 and other cytokine genes in tumors represents an alternative immunotherapeutic approach that may avoid systemic toxicity of recombinant cytokines. To achieve this goal of localizing gene expression, a heat-inducible adenoviral gene therapy vector with feline IL-12 being placed under the control of a heat inducible promoter was developed. The rationale for using this heat shock promoter (hsp70B) is that hyperthermia is most likely to be used as an adjuvant therapy with radiation and chemotherapy in the treatment of cancers. Heating the tumor leads to activation of the hsp promoter and subsequent local IL12 production. Materials/Methods: Cats with spontaneously arising soft tissue sarcomas presenting for treatment to the Colorado State University- Voss Veterinary Teaching Hospital and the Veterinary Teaching Hospital, NCSU were recruited for a Phase I dose escalation gene therapy trial. The cats underwent radiation therapy to a total dose of 48 Gy in 16 fractions. Four- five days later the gene construct was injected intratumorally. 24 hours post- injection the tumor was heated to a target temperature of 41oC for 60 minutes using a microwave applicator. Cats were treated at increasing dose levels of viral construct. Tumor expression of cytokines interleukin 12 and interferon- ? was quantitatively determined using Real time PCR. The cats were monitored for hematologic and hepatic toxicity. Results: High intratumoral levels of interleukin 12 were achieved with very low or absent interferon- ?. This could possibly be attributed to 1)In the large tumors there were low levels of IL- 12 due to low multiplicity of infection;2)Even with high levels of IL 12 there were not enough of IL- 12 target cells in the tumor (T- cells, NK cells) post- radiation therapy, to produce IFN?;3)Inhomogeneous heat distribution in the tumor, some parts were heated to 45°C- 46°C, resulting in direct hyperthermia induced cytotoxocity and these cells not able to produce IL 12. Hematologic and hepatic toxicity were limited and acceptable. Conclusions: It is possible to limit the toxicity of interleukin- 12 using the hyperthermia induced gene- therapy approach. Cats with spontaneously arising soft tissue sarcomas serve as an excellent model for the study of this approach which can be similarly applied to human cancer treatment.
2404
High Dose Rate Brachytherapy Boost in the Treatment of Extremity Soft-Tissue Sarcomas
O. E. Streeter,1 E. Chung,1 S. Chawla,4 G. Jozsef,1 M. A. Astrahan,1 S. Groshen,3 L. Menendez2 Radiation Oncology, Keck School of Medicine, University of Southern California, Los Angeles, CA, 2Orthopaedics, Keck School of Medicine, University of Southern California, Los Angeles, CA, 3Preventive Medicine, Keck School of Medicine, University of Southern California, Los Angeles, CA, 4Medicine, Keck School of Medicine, University of Southern California, Los Angeles, CA
1
Purpose/Objective: Retrospective review of our experience with high dose rate brachytherapy (HDR-BT) as a radiation boost following limb-sparing resection for extremity soft tissue sarcomas with a median follow-up of five years. We examined toxicity, local recurrence, metastatic disease development, and survival. Materials/Methods: Twelve patients were treated between April 1994 and September 1995 for extremity sarcomas. There were nine men and three women ranging in age from 31 to 73 years old (median 60 years). All were considered candidates for limb-sparing surgery prior to the start of any therapy. Six had malignant fibrous histiocytoma (50%), two had liposarcoma (17%), and one patient each had a leiomyosarcoma, synovial sarcoma, malignant schwannoma or spindle cell sarcoma (33%). Ten patients had high-grade tumors and two patients had low-grade tumors on initial biopsy. The sites of tumor origin were 8 in the lower extremity, 1 in the buttock, and 2 in the upper extremity; 1 patient had a lesion in both the upper and lower extremities. The treatment consisted of interstitial high dose rate brachytherapy (HDR-BT) following neoadjuvant chemotherapy in 8 of the 12 patients. Ten received external beam radiation therapy (EBRT), post operative radiation in one. The eight patients who received chemotherapy received either Ifosfamide/Mesna with or without Adriamycin, or MAID for 3 to 4 cycles. One patient also received Cisplatinum. Preoperative external beam radiation doses ranged from 4000 cGy to 5940 cGy, and were given in daily fractions of 180 cGy to 200 cGy. At the time of en bloc resection of the sarcoma, interstitial catheters were sutured into the surgical bed for remote afterloading HDR-BT delivery. The HDR-BT was administered using a HDR remote afterloading machine. Doses ranging from 1300 to 3000 cGy were delivered to the surgical tumor bed at a depth of 5 to 7.5 mm from the center of the Ir192 high dose rate source. Results: Eleven of the twelve patients experienced treatment-related toxicities. The toxicities were delayed wound healing in one, four required grafts or flaps, two experienced neurologic impairment, and four patients had chemotherapy-related neutropenia. With a
S577
S578
I. J. Radiation Oncology
● Biology ● Physics
Volume 60, Number 1, Supplement, 2004
median follow-up of 63 months, four patients have recurred. One patient developed a local recurrence at 37 months; two patients developed distant metastases (at 5 and 12 months) and one patient recurred with both local and distant disease at 24 months. The two patients with the very early metastases have died. The estimated five-year probability of surviving for the entire group was 0.83 ⫾ 0.17 (s.e.). The five-year probability of recurring was 0.33 ⫾ 0.14 (s.e.). Of the ten patients with high-grade sarcoma, seven received adjuvant chemotherapy and of these, two patients subsequently developed distant metastases. There were three patients with high-grade sarcoma who did not receive adjuvant chemotherapy, and one developed distant metastases. Conclusions: HDR-IBT appears to assist in local control of high-risk soft tissue sarcomas of the extremity following neoadjuvant therapy and en bloc resection with manageable toxicity. The local failure rate of 20% in high-grade lesions is comparable to the 19% local failure rate reported by Pister et al from Memorial Sloan-Kettering Cancer Center1. Neoadjuvant chemotherapy may contribute to local control as well as in overall survival 2. References: 1. Pisters PWT, Harrison LB, Leung DHY, et al. Long term results of a prospective randomized trial of adjuvant brachytherapy in soft tissue sarcoma. J Clin Oncol 14:859 – 868, 1996 2. DeLaney TF, Spiro IJ, Suit HD, Gebhardt MC, Hornicek FJ, et al. Neoadjuvant chemotherapy and radiotherapy for large extremity soft tissue sarcoma. Int. J. Radiation Oncology Biol. Phys., 56(4):1117–27, 2003
2405
Novel Feature Based Image Registration for Automated Patient Positioning
S. Samant,1 A. Parra,2 J. Wu3 SJRCH, Memphis, TN, 2University of Memphis, Memphis, TN, 3University of Tennessee, Memphis, TN
1
Purpose/Objective: In radiotherapy, current feature based image registration techniques (point or curve based) lack robustness and automation: curve based registrations require user input and can be affected by edge selection and noise, while point based techniques require equivalent and/or ordered point sets in the target and reference images. This limits their use for patient positioning using EPID imaging, and in optical based camera systems (bite-block intra-cranial positioning, external marker based body positioning and patient surface imaging). Here a branch-and-bound registration is presented that overcomes these problems. It is a faster alternative to mutual information (MI) registration, and more robust in marker based registration compared to singular value decomposition (SVD). Materials/Methods: The branch-and-bound (BB) algorithm presented here divides a bounded parameter space of rigid-body transformations in hyper-rectangles and evaluates a given transformation using either partial Hausdorff distance [1] or bipartite matching. It utilizes an exhaustive search based on user specified ranges. BB does not require the establishment of correspondence between the point sets, and is robust to outliers. Synthetic and actual anthropomorphic phantom data were used to test the algorithm. Images were acquired on a GE Discovery LightSpeed CT, Siemens BEAMVIEW video-EPID (VEPID), and a Varian PortalVision aS500 flat-panel. The VEPID images were corrected for optical spatial distortion, and flat-panel images were in-air flat fielded to correct for transistor response variations. Curves were automatically generated using generic edge detection algorithms. User modification of curves or tuning of edge detection was not allowed. External marker registrations (4mm lead markers placed externally) were carried out with unordered unequal point sets. Results: Using BB, marker based registrations (up to 7 markers) had an accuracy of ⬍0.1mm and ⬍0.1deg (CPU time ⫽ 0.1sec on 1.4GHz PC). For feature based registrations with BB, EPID-EPID (1400 points each) had an accuracy of 0.2mm (1 pixel ⫽ 0.5mm) and 0.1deg (CPU time ⫽ 4.5sec). DRR-EPID (2200 points for DRRs) had an accuracy of 1.1mm and 0.9deg (CPU time ⫽ 10.3sec). For MI with only in-air flat fielding, EPID-EPID had an accuracy ⬍0.5mm and DRR-EPID had an accuracy of 2– 4mm (CPU time ⫽ 90sec). For MI with additional flat fielding to account for beam quality changes, EPID-DRR had an accuracy of 1mm, with no change in the EPID-EPID accuracy.@½pick;z2d1110431370002;0;1 Conclusions: The BB algorithm yielded intra and inter-modality 2D registrations with comparable accuracy and faster times than MI, and did not require additional flat fielding or other processing of the EPID images as in the case of MI, making BB clinically simpler to use. It was also accurate and more robust for external marker registrations compared to SVD. Although this algorithm was tested on x-ray and synthetic data, it is readily applicable to a host of positioning systems in radiotherapy, including optical camera systems (bite-block intra-cranial positioning, external marker based body positioning and patient surface imaging) where outliers, and unequal and/ or unordered point sets are present. [1]Huttenlocher DP and Rucklidge WJ, A multi-resolution technique for comparing images using the Hausdorff distance Proc Third IEEE Conf Computer Vision and Pattern Recognition 705–706, 1993.
Conclusions: Brachytherapy provided a high level of local control for small tumours. Carcinologic, functional and cosmetic results were very good when implants were used as the exclusive treatment. Brachytherapy is a therapeutic alternative for peri-orificial face skin carcinomas because of the irregularity of contours and the difficulty in surgical reconstruction in this areas.
2403
A Phase I Hyperthermia- Induced Interleukin- 12 Gene- Therapy Trial in a Spontaneously Arising Feline Soft Tissue Sarcoma Model
F. Siddiqui,1 P. R. Avery,2 S. M. LaRue,1 M. L. Hauck,3 M. W. Dewhirst,4 C. Li,4 R. L. Ullrich1 Environmental and Radiological Health Sciences, Colorado State University, Fort Collins, CO, 2Department of Microbiology, Immunology and Pathology, Colorado State University, Fort Collins, CO, 3Department of Clinical Sciences, College of Veterinary Medicine, Raleigh, NC, 4Department of Radiation Oncology, Duke University, Durham, NC 1
Purpose/Objective: Cytokine genes are the most widely and extensively studied immunostimulatory agents in cancer gene therapy. In several studies, interleukin-12 (IL-12) was the most effective cytokine in inducing the eradication of experimental tumors, preventing development of metastases, and eliciting long-term antitumor immunity. Depending on the tumor model, IL-12 can exert anti- tumor activities via T cells, NK cells, or NKT cells. Induction of cytokines, such as IFN-g and IFN-inducible protein-l0, has also been implicated as a mechanism of antitumor activity of IL-12. Local and systemic administration of IL-12 protein has been studied in various murine models and in Phase I/II human trials. However, IL-12 protein therapy has been limited by dose-dependent toxicity. Local and efficient expression of IL-12 and other cytokine genes in tumors represents an alternative immunotherapeutic approach that may avoid systemic toxicity of recombinant cytokines. To achieve this goal of localizing gene expression, a heat-inducible adenoviral gene therapy vector with feline IL-12 being placed under the control of a heat inducible promoter was developed. The rationale for using this heat shock promoter (hsp70B) is that hyperthermia is most likely to be used as an adjuvant therapy with radiation and chemotherapy in the treatment of cancers. Heating the tumor leads to activation of the hsp promoter and subsequent local IL12 production. Materials/Methods: Cats with spontaneously arising soft tissue sarcomas presenting for treatment to the Colorado State University- Voss Veterinary Teaching Hospital and the Veterinary Teaching Hospital, NCSU were recruited for a Phase I dose escalation gene therapy trial. The cats underwent radiation therapy to a total dose of 48 Gy in 16 fractions. Four- five days later the gene construct was injected intratumorally. 24 hours post- injection the tumor was heated to a target temperature of 41oC for 60 minutes using a microwave applicator. Cats were treated at increasing dose levels of viral construct. Tumor expression of cytokines interleukin 12 and interferon- ? was quantitatively determined using Real time PCR. The cats were monitored for hematologic and hepatic toxicity. Results: High intratumoral levels of interleukin 12 were achieved with very low or absent interferon- ?. This could possibly be attributed to 1)In the large tumors there were low levels of IL- 12 due to low multiplicity of infection;2)Even with high levels of IL 12 there were not enough of IL- 12 target cells in the tumor (T- cells, NK cells) post- radiation therapy, to produce IFN?;3)Inhomogeneous heat distribution in the tumor, some parts were heated to 45°C- 46°C, resulting in direct hyperthermia induced cytotoxocity and these cells not able to produce IL 12. Hematologic and hepatic toxicity were limited and acceptable. Conclusions: It is possible to limit the toxicity of interleukin- 12 using the hyperthermia induced gene- therapy approach. Cats with spontaneously arising soft tissue sarcomas serve as an excellent model for the study of this approach which can be similarly applied to human cancer treatment.
2404
High Dose Rate Brachytherapy Boost in the Treatment of Extremity Soft-Tissue Sarcomas
O. E. Streeter,1 E. Chung,1 S. Chawla,4 G. Jozsef,1 M. A. Astrahan,1 S. Groshen,3 L. Menendez2 Radiation Oncology, Keck School of Medicine, University of Southern California, Los Angeles, CA, 2Orthopaedics, Keck School of Medicine, University of Southern California, Los Angeles, CA, 3Preventive Medicine, Keck School of Medicine, University of Southern California, Los Angeles, CA, 4Medicine, Keck School of Medicine, University of Southern California, Los Angeles, CA
1
Purpose/Objective: Retrospective review of our experience with high dose rate brachytherapy (HDR-BT) as a radiation boost following limb-sparing resection for extremity soft tissue sarcomas with a median follow-up of five years. We examined toxicity, local recurrence, metastatic disease development, and survival. Materials/Methods: Twelve patients were treated between April 1994 and September 1995 for extremity sarcomas. There were nine men and three women ranging in age from 31 to 73 years old (median 60 years). All were considered candidates for limb-sparing surgery prior to the start of any therapy. Six had malignant fibrous histiocytoma (50%), two had liposarcoma (17%), and one patient each had a leiomyosarcoma, synovial sarcoma, malignant schwannoma or spindle cell sarcoma (33%). Ten patients had high-grade tumors and two patients had low-grade tumors on initial biopsy. The sites of tumor origin were 8 in the lower extremity, 1 in the buttock, and 2 in the upper extremity; 1 patient had a lesion in both the upper and lower extremities. The treatment consisted of interstitial high dose rate brachytherapy (HDR-BT) following neoadjuvant chemotherapy in 8 of the 12 patients. Ten received external beam radiation therapy (EBRT), post operative radiation in one. The eight patients who received chemotherapy received either Ifosfamide/Mesna with or without Adriamycin, or MAID for 3 to 4 cycles. One patient also received Cisplatinum. Preoperative external beam radiation doses ranged from 4000 cGy to 5940 cGy, and were given in daily fractions of 180 cGy to 200 cGy. At the time of en bloc resection of the sarcoma, interstitial catheters were sutured into the surgical bed for remote afterloading HDR-BT delivery. The HDR-BT was administered using a HDR remote afterloading machine. Doses ranging from 1300 to 3000 cGy were delivered to the surgical tumor bed at a depth of 5 to 7.5 mm from the center of the Ir192 high dose rate source. Results: Eleven of the twelve patients experienced treatment-related toxicities. The toxicities were delayed wound healing in one, four required grafts or flaps, two experienced neurologic impairment, and four patients had chemotherapy-related neutropenia. With a
S577
S578
I. J. Radiation Oncology
● Biology ● Physics
Volume 60, Number 1, Supplement, 2004
median follow-up of 63 months, four patients have recurred. One patient developed a local recurrence at 37 months; two patients developed distant metastases (at 5 and 12 months) and one patient recurred with both local and distant disease at 24 months. The two patients with the very early metastases have died. The estimated five-year probability of surviving for the entire group was 0.83 ⫾ 0.17 (s.e.). The five-year probability of recurring was 0.33 ⫾ 0.14 (s.e.). Of the ten patients with high-grade sarcoma, seven received adjuvant chemotherapy and of these, two patients subsequently developed distant metastases. There were three patients with high-grade sarcoma who did not receive adjuvant chemotherapy, and one developed distant metastases. Conclusions: HDR-IBT appears to assist in local control of high-risk soft tissue sarcomas of the extremity following neoadjuvant therapy and en bloc resection with manageable toxicity. The local failure rate of 20% in high-grade lesions is comparable to the 19% local failure rate reported by Pister et al from Memorial Sloan-Kettering Cancer Center1. Neoadjuvant chemotherapy may contribute to local control as well as in overall survival 2. References: 1. Pisters PWT, Harrison LB, Leung DHY, et al. Long term results of a prospective randomized trial of adjuvant brachytherapy in soft tissue sarcoma. J Clin Oncol 14:859 – 868, 1996 2. DeLaney TF, Spiro IJ, Suit HD, Gebhardt MC, Hornicek FJ, et al. Neoadjuvant chemotherapy and radiotherapy for large extremity soft tissue sarcoma. Int. J. Radiation Oncology Biol. Phys., 56(4):1117–27, 2003
2405
Novel Feature Based Image Registration for Automated Patient Positioning
S. Samant,1 A. Parra,2 J. Wu3 SJRCH, Memphis, TN, 2University of Memphis, Memphis, TN, 3University of Tennessee, Memphis, TN
1
Purpose/Objective: In radiotherapy, current feature based image registration techniques (point or curve based) lack robustness and automation: curve based registrations require user input and can be affected by edge selection and noise, while point based techniques require equivalent and/or ordered point sets in the target and reference images. This limits their use for patient positioning using EPID imaging, and in optical based camera systems (bite-block intra-cranial positioning, external marker based body positioning and patient surface imaging). Here a branch-and-bound registration is presented that overcomes these problems. It is a faster alternative to mutual information (MI) registration, and more robust in marker based registration compared to singular value decomposition (SVD). Materials/Methods: The branch-and-bound (BB) algorithm presented here divides a bounded parameter space of rigid-body transformations in hyper-rectangles and evaluates a given transformation using either partial Hausdorff distance [1] or bipartite matching. It utilizes an exhaustive search based on user specified ranges. BB does not require the establishment of correspondence between the point sets, and is robust to outliers. Synthetic and actual anthropomorphic phantom data were used to test the algorithm. Images were acquired on a GE Discovery LightSpeed CT, Siemens BEAMVIEW video-EPID (VEPID), and a Varian PortalVision aS500 flat-panel. The VEPID images were corrected for optical spatial distortion, and flat-panel images were in-air flat fielded to correct for transistor response variations. Curves were automatically generated using generic edge detection algorithms. User modification of curves or tuning of edge detection was not allowed. External marker registrations (4mm lead markers placed externally) were carried out with unordered unequal point sets. Results: Using BB, marker based registrations (up to 7 markers) had an accuracy of ⬍0.1mm and ⬍0.1deg (CPU time ⫽ 0.1sec on 1.4GHz PC). For feature based registrations with BB, EPID-EPID (1400 points each) had an accuracy of 0.2mm (1 pixel ⫽ 0.5mm) and 0.1deg (CPU time ⫽ 4.5sec). DRR-EPID (2200 points for DRRs) had an accuracy of 1.1mm and 0.9deg (CPU time ⫽ 10.3sec). For MI with only in-air flat fielding, EPID-EPID had an accuracy ⬍0.5mm and DRR-EPID had an accuracy of 2– 4mm (CPU time ⫽ 90sec). For MI with additional flat fielding to account for beam quality changes, EPID-DRR had an accuracy of 1mm, with no change in the EPID-EPID accuracy.@½pick;z2d1110431370002;0;1 Conclusions: The BB algorithm yielded intra and inter-modality 2D registrations with comparable accuracy and faster times than MI, and did not require additional flat fielding or other processing of the EPID images as in the case of MI, making BB clinically simpler to use. It was also accurate and more robust for external marker registrations compared to SVD. Although this algorithm was tested on x-ray and synthetic data, it is readily applicable to a host of positioning systems in radiotherapy, including optical camera systems (bite-block intra-cranial positioning, external marker based body positioning and patient surface imaging) where outliers, and unequal and/ or unordered point sets are present. [1]Huttenlocher DP and Rucklidge WJ, A multi-resolution technique for comparing images using the Hausdorff distance Proc Third IEEE Conf Computer Vision and Pattern Recognition 705–706, 1993.