International Journal of
Radiation Oncology biology
physics
www.redjournal.org
Clinical Investigation
Spot Scanning Proton Therapy for Malignancies of the Base of Skull: Treatment Planning, Acute Toxicities, and Preliminary Clinical Outcomes David R. Grosshans, MD, PhD,* X. Ronald Zhu, PhD,y Adam Melancon, PhD,y Pamela K. Allen, PhD,* Falk Poenisch, PhD,y Matthew Palmer, CMD, MBA,y Mary Frances McAleer, MD, PhD,* Susan L. McGovern, MD, PhD,* Michael Gillin, PhD,y Franco DeMonte, MD,z Eric L. Chang, MD,x Paul D. Brown, MD,* and Anita Mahajan, MD* Departments of *Radiation Oncology, yRadiation Physics, and zNeurosurgery, The University of Texas MD Anderson Cancer Center, Houston, Texas; and xDepartment of Radiation Oncology, University of Southern California Keck School of Medicine, Los Angeles, California Received Feb 20, 2014, and in revised form Jun 30, 2014. Accepted for publication Jul 3, 2014.
Summary We describe here our early experience with the use of spot scanning proton therapy for chordoma or chondrosarcoma of the skull base. Comparisons of treatment planning techniques confirmed improvements in high-dose conformality from spot scanning versus passive scattering approaches, and early clinical outcomes suggest that single- or multifield optimized plans were well tolerated and produced good disease control. A simultaneous integrated boost technique may be advantageous.
Purpose: To describe treatment planning techniques and early clinical outcomes in patients treated with spot scanning proton therapy for chordoma or chondrosarcoma of the skull base. Methods and Materials: From June 2010 through August 2011, 15 patients were treated with spot scanning proton therapy for chordoma (nZ10) or chondrosarcoma (nZ5) at a single institution. Toxicity was prospectively evaluated and scored weekly and at all follow-up visits according to Common Terminology Criteria for Adverse Events, version 3.0. Treatment planning techniques and dosimetric data were recorded and compared with those of passive scattering plans created with clinically applicable dose constraints. Results: Ten patients were treated with single-field-optimized scanning beam plans and 5 with multifield-optimized intensity modulated proton therapy. All but 2 patients received a simultaneous integrated boost as well. The mean prescribed radiation doses were 69.8 Gy (relative biological effectiveness [RBE]; range, 68-70 Gy [RBE]) for chordoma and 68.4 Gy (RBE) (range, 66-70) for chondrosarcoma. In comparison with passive scattering plans, spot scanning plans demonstrated improved high-dose conformality and sparing of temporal lobes and brainstem. Clinically, the most common acute toxicities included fatigue (grade 2 for 2 patients, grade 1 for 8 patients) and nausea (grade 2 for 2 patients, grade 1 for 6 patients). No toxicities of grades 3 to 5 were recorded. At a median follow-up time of 27 months (range, 13-42 months), 1 patient had experienced local recurrence and a second developed distant metastatic
Reprint requests to: David R. Grosshans, MD, PhD, Division of Radiation Oncology, The University of Texas MD Anderson Cancer Center,
Int J Radiation Oncol Biol Phys, Vol. 90, No. 3, pp. 540e546, 2014 0360-3016/$ - see front matter Ó 2014 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.ijrobp.2014.07.005
1515 Holcombe Blvd, Unit 1150, Houston, TX 77030. Tel: (713) 7458795; E-mail: [email protected] Conflict of interest: none.
Volume 90 Number 3 2014
Spot scanning PT for skull base tumors
541
disease. Two patients had magnetic resonance imaging-documented temporal lobe changes, and a third patient developed facial numbness. No other subacute or late effects were recorded. Conclusions: In comparison to passive scattering, treatment plans for spot scanning proton therapy displayed improved high-dose conformality. Clinically, the treatment was well tolerated, and with short-term follow-up, disease control rates and toxicity profiles were favorable. Ó 2014 Elsevier Inc.
Introduction Chordomas and chondrosarcomas of the skull base are locally aggressive tumors. Although the risk of distant metastasis is low, without adequate therapy, these tumors often recur locally with significant morbidity and mortality (1). Particle therapy has shown promise for the treatment of intrinsic malignancies of the skull base. Of the types of particles used in the treatment of these malignancies, proton therapy (PT) has been used most widely (2-5). The physical properties of protons allow high doses to be delivered to target volumes that are often close to critical normal tissues. Historically, PT has been delivered most commonly by passive scattering techniques, in which customized metal apertures and acrylic compensators are used to shape the lateral and distal aspects of individual proton beams. In contrast, with spot scanning proton therapy, a small pristine proton beam is magnetically scanned to cover the lateral aspects of the target. The depth of the dose is controlled through use of different proton beam energies as well as range shifters. In contrast to passive scatter deliver, scanning beam offers greater control over the proximal aspects of the beam and potentially improved conformality of high-dose regions. Moreover, similar to advanced photon therapy techniques, scanning beam PT allows multiple target volumes to be treated to discrete doses through the use of a simultaneous integrated boost (SIB) technique. Worldwide, few centers have extensive experience with scanning beam PT for base-of-skull malignancies. To date, investigators at the Paul Scherrer Institute have the most indepth experience and remain the only group to publish long-term outcomes (6, 7). Here we report our initial clinical and treatment planning experience with spot scanning PT to treat chordomas and chondrosarcomas of the skull base. In contrast to previous reports, most patients in this study were treated with SIB technique or with fullcourse multifield optimized (MFO) intensity modulated proton therapy (IMPT).
Methods and Materials From June 2010 through August 2011, 15 patients were treated for base of skull chordomas or chondrosarcomas
using active spot scanning PT as part of an institutional review board-approved prospective protocol. For PT planning and treatment, all patients were immobilized with a thermoplastic mask and bite block to minimize rotational setup uncertainties. Target volumes and organs at risk (OARs) were delineated by a staff radiation oncologist experienced in the treatment of malignancies of the skull base. The gross tumor volume (GTV) encompassed any gross residual disease. Clinical target volumes (CTVs) encompassed adjacent areas at risk for recurrence as judged by the treating physician. In 13 patients, 2 separate CTVs were delineated: CTV1 was inclusive of the GTV with a small margin (typically 5-8 mm) added to cover areas at highest risk, such as areas of gross disease before surgical resection. CTV1 was the highest prescription target. CTV2 included additional expansions beyond CTV1 to cover areas at lower risk and the surgical pathway, if the treating physician felt that such areas could be safely covered. In the patient cohort reported here, critical normal tissues including the optic chiasm, brainstem, or temporal lobes in close proximity to target volumes. Thus, out of clinical necessity, in order to avoid overdosing of such structures, a relatively small planning target volume (PTV) of 2 to 3 mm was used for planning purposes for both target volumes. Smaller expansions such as these may be partially justified based on our documented setup uncertainties for head and neck and brain tumor patients as well as previous studies of tissue-specific range uncertainties (8). Regardless, it is important to note that in patients with greater distances between targets volumes and critical normal tissues, a larger expansion would be favored. Of equal importance, whereas the PTV concept may not be directly translated for use in proton therapy, because range uncertainties are beam direction-specific, such individual beam-specific planning structures are not available in our treatment planning system, nor is robust optimization available, which would be the preferred approach. An Eclipse treatment planning system (version 8.9; Varian Medical Systems, Palo Alto, CA) was used in all cases. Beam parameters have been previously described (9). Doses were prescribed in terms of Gy (relative biological effectiveness [RBE]) using a value of 1.1. Inverse planning with dose constraints both for target volumes and organs at risk was used for both single-field optimization (SFO) and MFO. In contrast to previous reports of scanning beam for base of skull tumors, in most cases, a
542
Grosshans et al.
International Journal of Radiation Oncology Biology Physics
simultaneous integrated boost was used in order to deliver differential doses to high- and low-risk target volumes as described above. Such an integrated boost technique is a natural extension of MFO. However, SFO with an integrated boost is also easily achieved with the scanning beam. In our experience, SFO plans with an integrated boost are typically of higher quality than sequential boost plans, in which each plan is optimized separately and then combined. Moreover, the number of quality assurance procedures is limited to one, using an integrated rather than sequential approach. Planning constraints emphasized maximum target coverage while attempting to limit the brainstem volume receiving 60 Gy (V60) to
Radiation Oncology biology
physics
www.redjournal.org
Clinical Investigation
Spot Scanning Proton Therapy for Malignancies of the Base of Skull: Treatment Planning, Acute Toxicities, and Preliminary Clinical Outcomes David R. Grosshans, MD, PhD,* X. Ronald Zhu, PhD,y Adam Melancon, PhD,y Pamela K. Allen, PhD,* Falk Poenisch, PhD,y Matthew Palmer, CMD, MBA,y Mary Frances McAleer, MD, PhD,* Susan L. McGovern, MD, PhD,* Michael Gillin, PhD,y Franco DeMonte, MD,z Eric L. Chang, MD,x Paul D. Brown, MD,* and Anita Mahajan, MD* Departments of *Radiation Oncology, yRadiation Physics, and zNeurosurgery, The University of Texas MD Anderson Cancer Center, Houston, Texas; and xDepartment of Radiation Oncology, University of Southern California Keck School of Medicine, Los Angeles, California Received Feb 20, 2014, and in revised form Jun 30, 2014. Accepted for publication Jul 3, 2014.
Summary We describe here our early experience with the use of spot scanning proton therapy for chordoma or chondrosarcoma of the skull base. Comparisons of treatment planning techniques confirmed improvements in high-dose conformality from spot scanning versus passive scattering approaches, and early clinical outcomes suggest that single- or multifield optimized plans were well tolerated and produced good disease control. A simultaneous integrated boost technique may be advantageous.
Purpose: To describe treatment planning techniques and early clinical outcomes in patients treated with spot scanning proton therapy for chordoma or chondrosarcoma of the skull base. Methods and Materials: From June 2010 through August 2011, 15 patients were treated with spot scanning proton therapy for chordoma (nZ10) or chondrosarcoma (nZ5) at a single institution. Toxicity was prospectively evaluated and scored weekly and at all follow-up visits according to Common Terminology Criteria for Adverse Events, version 3.0. Treatment planning techniques and dosimetric data were recorded and compared with those of passive scattering plans created with clinically applicable dose constraints. Results: Ten patients were treated with single-field-optimized scanning beam plans and 5 with multifield-optimized intensity modulated proton therapy. All but 2 patients received a simultaneous integrated boost as well. The mean prescribed radiation doses were 69.8 Gy (relative biological effectiveness [RBE]; range, 68-70 Gy [RBE]) for chordoma and 68.4 Gy (RBE) (range, 66-70) for chondrosarcoma. In comparison with passive scattering plans, spot scanning plans demonstrated improved high-dose conformality and sparing of temporal lobes and brainstem. Clinically, the most common acute toxicities included fatigue (grade 2 for 2 patients, grade 1 for 8 patients) and nausea (grade 2 for 2 patients, grade 1 for 6 patients). No toxicities of grades 3 to 5 were recorded. At a median follow-up time of 27 months (range, 13-42 months), 1 patient had experienced local recurrence and a second developed distant metastatic
Reprint requests to: David R. Grosshans, MD, PhD, Division of Radiation Oncology, The University of Texas MD Anderson Cancer Center,
Int J Radiation Oncol Biol Phys, Vol. 90, No. 3, pp. 540e546, 2014 0360-3016/$ - see front matter Ó 2014 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.ijrobp.2014.07.005
1515 Holcombe Blvd, Unit 1150, Houston, TX 77030. Tel: (713) 7458795; E-mail: [email protected] Conflict of interest: none.
Volume 90 Number 3 2014
Spot scanning PT for skull base tumors
541
disease. Two patients had magnetic resonance imaging-documented temporal lobe changes, and a third patient developed facial numbness. No other subacute or late effects were recorded. Conclusions: In comparison to passive scattering, treatment plans for spot scanning proton therapy displayed improved high-dose conformality. Clinically, the treatment was well tolerated, and with short-term follow-up, disease control rates and toxicity profiles were favorable. Ó 2014 Elsevier Inc.
Introduction Chordomas and chondrosarcomas of the skull base are locally aggressive tumors. Although the risk of distant metastasis is low, without adequate therapy, these tumors often recur locally with significant morbidity and mortality (1). Particle therapy has shown promise for the treatment of intrinsic malignancies of the skull base. Of the types of particles used in the treatment of these malignancies, proton therapy (PT) has been used most widely (2-5). The physical properties of protons allow high doses to be delivered to target volumes that are often close to critical normal tissues. Historically, PT has been delivered most commonly by passive scattering techniques, in which customized metal apertures and acrylic compensators are used to shape the lateral and distal aspects of individual proton beams. In contrast, with spot scanning proton therapy, a small pristine proton beam is magnetically scanned to cover the lateral aspects of the target. The depth of the dose is controlled through use of different proton beam energies as well as range shifters. In contrast to passive scatter deliver, scanning beam offers greater control over the proximal aspects of the beam and potentially improved conformality of high-dose regions. Moreover, similar to advanced photon therapy techniques, scanning beam PT allows multiple target volumes to be treated to discrete doses through the use of a simultaneous integrated boost (SIB) technique. Worldwide, few centers have extensive experience with scanning beam PT for base-of-skull malignancies. To date, investigators at the Paul Scherrer Institute have the most indepth experience and remain the only group to publish long-term outcomes (6, 7). Here we report our initial clinical and treatment planning experience with spot scanning PT to treat chordomas and chondrosarcomas of the skull base. In contrast to previous reports, most patients in this study were treated with SIB technique or with fullcourse multifield optimized (MFO) intensity modulated proton therapy (IMPT).
Methods and Materials From June 2010 through August 2011, 15 patients were treated for base of skull chordomas or chondrosarcomas
using active spot scanning PT as part of an institutional review board-approved prospective protocol. For PT planning and treatment, all patients were immobilized with a thermoplastic mask and bite block to minimize rotational setup uncertainties. Target volumes and organs at risk (OARs) were delineated by a staff radiation oncologist experienced in the treatment of malignancies of the skull base. The gross tumor volume (GTV) encompassed any gross residual disease. Clinical target volumes (CTVs) encompassed adjacent areas at risk for recurrence as judged by the treating physician. In 13 patients, 2 separate CTVs were delineated: CTV1 was inclusive of the GTV with a small margin (typically 5-8 mm) added to cover areas at highest risk, such as areas of gross disease before surgical resection. CTV1 was the highest prescription target. CTV2 included additional expansions beyond CTV1 to cover areas at lower risk and the surgical pathway, if the treating physician felt that such areas could be safely covered. In the patient cohort reported here, critical normal tissues including the optic chiasm, brainstem, or temporal lobes in close proximity to target volumes. Thus, out of clinical necessity, in order to avoid overdosing of such structures, a relatively small planning target volume (PTV) of 2 to 3 mm was used for planning purposes for both target volumes. Smaller expansions such as these may be partially justified based on our documented setup uncertainties for head and neck and brain tumor patients as well as previous studies of tissue-specific range uncertainties (8). Regardless, it is important to note that in patients with greater distances between targets volumes and critical normal tissues, a larger expansion would be favored. Of equal importance, whereas the PTV concept may not be directly translated for use in proton therapy, because range uncertainties are beam direction-specific, such individual beam-specific planning structures are not available in our treatment planning system, nor is robust optimization available, which would be the preferred approach. An Eclipse treatment planning system (version 8.9; Varian Medical Systems, Palo Alto, CA) was used in all cases. Beam parameters have been previously described (9). Doses were prescribed in terms of Gy (relative biological effectiveness [RBE]) using a value of 1.1. Inverse planning with dose constraints both for target volumes and organs at risk was used for both single-field optimization (SFO) and MFO. In contrast to previous reports of scanning beam for base of skull tumors, in most cases, a
542
Grosshans et al.
International Journal of Radiation Oncology Biology Physics
simultaneous integrated boost was used in order to deliver differential doses to high- and low-risk target volumes as described above. Such an integrated boost technique is a natural extension of MFO. However, SFO with an integrated boost is also easily achieved with the scanning beam. In our experience, SFO plans with an integrated boost are typically of higher quality than sequential boost plans, in which each plan is optimized separately and then combined. Moreover, the number of quality assurance procedures is limited to one, using an integrated rather than sequential approach. Planning constraints emphasized maximum target coverage while attempting to limit the brainstem volume receiving 60 Gy (V60) to
