Clinical Proton Radiation Therapy Research at the Francis H. Burr Proton Therapy Center

Transcription

1 Technology in Cancer Research and Treatment ISSN Volume 6, Number 4 Supplement, August 2007 Adenine Press (2007) Clinical Proton Radiation Therapy Research at the Francis H. Burr Proton Therapy Center The Francis H. Burr Proton Therapy Center has a 230 MeV cyclotron from which proton beams are directed to two isocentric gantries, a stereotactic intracranial beam line, and an eye line. Because of improved physical dose distribution, proton radiotherapy allows dose escalation to improve local tumor control in anatomic sites and histologies where local control is suboptimal with photons. The improved dose localization also reduces normaltissue doses with an anticipated reduction in acute and late toxicity. Clinical treatment protocols, developed to exploit the dosimetric advantages of protons over photons, have been grouped into two broad categories. In the first, dose is escalated for anatomic sites where local control with conventional radiation doses has been suboptimal. In the second, normal-tissue sparing with protons is designed to minimize acute and late toxicity. Treatment of patients on clinical research protocols has been encouraged. Patient treatments began on the first gantry in November 2001; on the eye line in April 2002; on the second gantry in May 2002; and on the stereotactic intracranial line in August The facility currently treats 60 patients per day, including up to six children daily under anesthesia. Dose-escalation studies have been completed for early stage prostate cancer (in conjunction with Loma Linda University) and sarcomas of the cervical spine/base of skull and thoracolumbosacral spine. Protocols are in progress or development for carcinoma of the nasopharynx, paranasal sinus carcinoma, non-small-cell lung carcinoma, locally advanced carcinoma of the prostate, hepatocellular carcinoma, and pancreatic cancer. Studies evaluating the use of protons for morbidity reduction include protocols for craniospinal irradiation in conjunction with systemic chemotherapy for medulloblastoma, retinoblastoma, pediatric rhabdomyosarcoma, other pediatric sarcomas, and accelerated, hypofractionated partial breast irradiation for T1N0 breast carcinomas. For pediatric patients, protons have also been accepted as an alternative to photons for children enrolled in Children s Oncology Group (COG) protocols. Treatment of patients on these studies has often required the development of new treatment techniques (i.e., matching abutting fields for craniospinal irradiation), respiratory gating, and development of appropriate clinical infrastructure support (i.e., increase in availability of pediatric anesthesia) to allow appropriate treatment. In addition, a clinical research infrastructure for protocol development and data management is required. Results to date indicate that proton radiation therapy offers several potential treatment advantages to patients that can be studied in the setting of clinical trials. Patients willingness to enter these clinical trials seems to be quite high; accrual to selected studies has been favorable. Thomas F. DeLaney, M.D. Division of Radiation Oncology Massachusetts General Hospital Cancer Center 55 Fruit Street Boston, MA 02114, USA Key words: Proton radiation therapy; and Clinical operations. Introduction The proton center at Massachusetts General Hospital (MGH), the Francis H. Burr Proton Therapy Center (FHBPTC), was named for Francis Burr, or Hooks Burr, as he was known, former chairman of the board of MGH. Mr. Burr was Corresponding Author: Thomas F. DeLaney, M.D. tdelaney@partners.org 61

2 62 DeLaney a major advocate for building this facility, once Herman D. Suit M.D., D. Phil. had convinced him that the proton beam was a better beam. The FHBPTC was built with approximately 40 percent financing from National Cancer Institute (NCI) grants and the remainder with charitable contributions and contributions from MGH. It was particularly important to have an advocate on the board who was able to help us realize the vision. From a clinician s point of view, a facility such as this is absolutely magnificent. To be able to control a beam by using the Bragg peak and eliminating exit dose is very important in radiation treatment. Our facility has a 230-MeV cyclotron, built by IBA, of Belgium, from which we steer beam into two gantries and a fixed-beam room that has two horizontal beams. Gantry rooms are equipped with six-degree-of-freedom patient positioning devices. Components in the facility were commissioned serially over time. Our first treatment was delivered in November of 2001, in the first gantry; the second gantry was partially commissioned for prostate treatments in 2003 and then, in January 2004, for all sites. The eye line and the eye treatment program, previously based at the Harvard cyclotron, was transferred in April of 2002, and a stereotactic radiosurgery/radiotherapy device was commissioned in July This experience is typical for hospital-based proton facilities, and reflects the fact that a limited number of physics personnel are available to effect commissioning. The number of patients treated has risen every year (Table I); we are at present treating 55 patients between the two gantries during a 10-hour treatment day. We plan to go to a longer treatment day, and are balancing that with some ongoing physics developments currently being worked on after hours. Table I Patients Treated Since Inception, FHBPTC. Year Number Treated 1 (11/01 10/02) (11/02 10/03) (11/03 10/04) (11/04 10/05) 5 (11/05 10/06, estimate) Clinical Operations and Research Interests At our facility, as I think as representatives of any of the facilities will attest, one is always trying to improve efficiency. Once personnel begin to use a proton facility, they may realize that physical renovations are necessary to address the needs of the patient population. In our case at MGH, we have a large pediatric population. We need to increase the pediatric anesthesia recovery space; currently we are able to treat six children per day under anesthesia, but are hoping to be able to treat eight in the near future. In another instance, we have recently commissioned a stereotactic device that was essentially upgraded from the original neurosurgical device employed at the Harvard Cyclotron Laboratory (HCL). Current projections are that we will be treating three radiosurgery cases per week, and then will have at least five patients who will be undergoing fractionated stereotactic radiotherapy weekly with this service. The beam line for this was developed in conjunction with Jay Flanz, Ph.D., and his group, and Hanne Kooy, Ph. D., upgraded the treatment planning system for this. Essentially the patient will be on a couch that has 30 degrees of tilt in either direction, and it rotates as well, to offer a number of potential beam entry angles from a horizontal fixed beam. Because it is a single-scattered beam with an 8-centimeter field diameter, it has a very sharp penumbra which is quite suitable for intracranial uses. We are also looking into using a multi-leaf collimator. In fiscal year 2006, 84% of our patients were adults; 16% were children. Nearly 60% of patients are treated on the gantries; 27% are eye patients; and 14% are patients undergoing fractionated or primarily single fraction stereotactic radiosurgery treatments. We project nearly 12,000 treatments for the 2006 fiscal year, 93% of which are given on the gantry; this reflects the fact that eye treatments are generally hypofractionated treatments (five fractions per patient) and that the stereotactic treatments have been, for the most part, single-fraction radiosurgery. Over 25% of our fractions are administered to pediatric patients, about half of whom are under anesthesia and about half of whom are awake children who are over the age of five years. A breakdown of the patient population treated on the gantries reveals a distribution that reflects a combination of clinical indications, clinical expertise, and, in our institution, research interests. We are most interested in studying patients for whom protons seem a good fit (Table II); certainly, intracranial tumors and skull-based tumors are ones for which there are a number of indications for the use of protons because of the optic structures nearby, and the issue of brain stem tolerance is significant. However, we also have a large sarcoma group, and we treat many patients with spinal tumors; in both these instances, the dose distribution of protons conveys a significant therapeutic advantage. We recently opened a nasopharynx protocol and anticipate treating more patients with nasopharyngeal tumors; we also expect to treat some with hepatic lesions and a handful of patients having truncal and extremity sarcomas. To date, our experience with lung cancer treatment with protons is modest; we have been able to use respiratory gating for the last year, but primarily we are using protons for patients who are medically inoperable and have early-stage disease. That is a relatively small group of the overall lung-cancer population. Our experience does not imply that one could not use proton therapy for patients requir-

3 Protons at the Francis H. Burr Center 63 Table II FHBPTC Patient Population (Treated on Gantries). Site Brain Spine Prostate Skull base Head & neck Trunk and ext. sarcoma GI Lung Percent of patients ing boosting of the primary site, and for patients with bulky Stage III disease; we simply have not done that yet. In terms of our clinical research objectives, we seek to improve local control with dose escalation and to expand the range of tumor sites and types that have been treated historically at the Harvard Cyclotron Laboratory. A related objective is to carefully quantify, where possible, normal-tissue dose response. We are interested in reducing both acute and late treatment-related morbidity, and are investigating one special area of interest: in the combined modality therapy era, where treatment breaks are often required because of acute mucositis or problems with blood counts, we think that by reducing the amount of mucosa in the field with protons and reducing the amount of marrow in the field, that one can improve the intensity of both radiotherapy and chemotherapy. One of the great areas of promise associated with protons is quality of life. The ability to avoid radiation exposure of normal tissues has obvious implications for this issue: Dr. Herman Suit, one of the pioneers of proton radiation therapy, is fond of saying, side effects do not occur in unirradiated tissues. To measure this aspect in some way, we are doing formal quality-of-life studies, particularly in pediatric patients but also in our patients with pelvic malignancies. Dr. James Talcott, one of the medical oncologists in our group, has a major interest in quality of life and has developed some instruments, particularly for use of patients with prostate cancer. Overview of Proton Clinical Research at FHBPTC Proton clinical research is spread out across our department. Many radiation oncologists and specialists from other disciplines collaborate. At present we have protocols either opened, completed, or under development in all of the clinical areas with the exception of gynecology, and that probably will change shortly. The common denominator in all these efforts is the focus on research; we are very interested in exploiting the advantages of protons in whatever anatomic sites wherein those advantages seem most apparent. The possibilities are almost legion. We have a long-standing interest in prostate cancer, and have participated in collaborative studies with the group from Loma Linda. Nancy Tarbell, M.D., Torunn Yock, M.D., and, more recently, Yen- Lin Chen, M.D. are heading up our pediatric effort. Jay Loeffler, M.D. and several others are doing central nervous system (CNS) tumors. We have active head and neck protocols. Ted Hong, M.D. recently joined us and is carrying on a liver protocol; he also has interesting ideas for both pancreas and rectal carcinomas. Herman Suit M.D., D.Phil. built a very active sarcoma program, and myself and David Kirsch, M.D., Ph.D. are carrying on that tradition. John Munzenrider, M.D. is heading up our eye effort; Noah Choi, M.D., our thoracic program. We have done a study and published a partial breast radiation technique experience (1). Among our clinical studies are several open protocols. We have a open protocol for patients with carcinoma of the nasopharynx, essentially building upon the Intergroup experience but substituting protons for photons with the aim of reducing interruptions in treatment and delivering more chemotherapy. We have completed a large spine sarcoma protocol that will be presented at the meeting of the American Society for Therapeutic Radiology and Oncology (ASTRO) this year, and have a base-of-skull randomized sarcoma experience that initially started in the Proton Radiation Oncology Group (PROG) and is now completed. We have reviewed the patients previously treated at the Harvard Cyclotron Laboratory for paranasal sinus carcinomas; that experience has been accepted for publication. For patients with locally advanced prostate cancer, we have a proposal that has been accepted for a protocol by the Radiation Therapy Oncology Group (RTOG); it will be, essentially, the first RTOG protocol in which protons will be used, and it features neo-adjuvant hormones and protons for treatment of patients with locally advanced prostate cancer. We are completing a Phase I study of hepatocellular carcinoma treated with escalated doses, based upon the EUD of the normal liver that is irradiated. We have proposals from Dr. Hong for a hypofractionated preoperative chemoradiation protocol for rectal cancer, essentially using the 5 Gray times 5 protocol that has been used in Europe but combining that with 5-fluorouracil (5-FU). We think that is something one could do with protons that one would be hard pressed to do with photons; we similarly wish to use the same approach for patients with cancer of the pancreas: essentially give a one-week rapid preoperative course of chemoradiation, followed shortly thereafter by surgery. This concept is interesting from a radiobiology point of view: if one studies the radiosensitization curves with 5-FU, there is substantially more radiosensitization at 4 or 5 Gray than there is at 2 Gray, so one might get more bang for the buck out of the equivalent dose of 5-FU. In lung cancer, which is an area that we expect to expand greatly in the future, Dr. Choi has proposed a protocol for

4 64 DeLaney hypofractionated treatment of medically inoperable disease. The essential objective is to look at reoxygenation of the tumor using 2-deoxy-2-[F-18]fluoro-D:-glucose with positron emission tomography (FDG-PET). Discussion Clinical research protocols at FHBPTC are opportunities to analyze in depth the uses of proton radiation and to assess its role in relation to other modalities, radiotherapeutic and nonradiotherapeutic. Our belief is that protons will have manifold clinical applications; our protocols are designed to identify potential applications and to evaluate the performance of proton radiation when it is employed in them. Nasopharyngeal Carcinoma Current experience in treating patients with locally advanced nasopharyngeal carcinomas with skull base invasion indicates that local recurrences might occur in about a third of them. This protocol is designed to use a combination of protons and intensity-modulated radiation therapy (IMRT) photons; we expect that about 20 Gray of the total dose will be given with IMRT, the remainder with protons, delivering a total dose of 70 cobalt gray equivalent (CGE) to the primary site. (It is worth noting that, to do comparative evaluations of protons and IMRT, one would wish to do IMRT and proton planning on the same platform. If proton planning systems are incompatible with photon versions, interpretation of results can be confounded.) In our protocol, the nodal volumes will be encompassed with shrinking fields. This protocol is currently open, and it involves concurrent cisplatinum with 100 milligrams per meter squared given every three weeks, to be followed by adjuvant 5-FU and cisplatinum after completion of radiotherapy. In the Intergroup experience, just over half of the patients received all of the adjuvant chemotherapy, so we anticipate that we should see fewer treatment breaks during the concurrent platinum and we can quantify how much additional postoperative adjuvant platinum is given. This study also gives us the opportunity to do IMRT versus proton dose comparisons. Paranasal Sinus We begin with data from the Harvard Cyclotron Laboratory experience. In this study of 96 patients, the three-year local control rate was 86 percent, which is about 30 percent higher than what was historically expected with photons. This outcome was achieved with a median dose of 71 Gray. The paranasal sinuses represent a very challenging volume to treat with photons because of the optic structures nearby. Ultimate local control, including surgical salvage, was 90 percent. When one looks at disease-free and overall survival, however, about half the patients are failing distantly. The current proposal, then, is to begin to treat these patients with concurrent chemotherapy to essentially begin to address the issue of distant metastases in these patients. Cancer of the Prostate We use a technique similar to that of the Loma Linda facility: treating a single lateral field each day. Our patients currently are being treated with B-mode acquisition and targeting (BAT) ultrasound guidance. We have put fiducial markers in some patients, the experience of which, however, can be uncomfortable for some patients. I would add, also that some physicians are not comfortable putting them in. Hence the use of BAT ultrasound. In one of our protocols, a cooperative study performed by Loma Linda and MGH, patients were randomized to receive a total dose of either 70 or 79 Gray (2). In both low-risk and intermediate/high-risk patients, there was a significant gain in biochemical disease-free survival. In the overall group, the risk of biochemical failure was reduced from about 35% to 18% with a 9 Gray increase in dose. Just from a radiobiology point of view, this is interesting. One can probably do a gamma factor calculation based upon that outcome: the gamma factor was about 2, i.e., for every Gray of increase in dose, the local control rate increased about 2%. That certainly reinforces our bias that there is a dose response and that with higher doses, increased local control should supervene. The other important feature of this study was that there was no difference in grade 3 morbidity between patients treated with 70 or 79 Gray, a testament to the fact that protons were used for dose escalation. Another cooperative effort between our group and Loma Linda involves a study in which patients with early stage disease were treated with protons alone just to the prostate and seminal vesicles to 50 CGE, followed by a boost to the prostate of another 32 CGE. This study accrued 84 patients and completed accrual in about a year and a half. It is still too early to analyze these data. MGH investigators, Drs. Shipley, Zietman,and Coen have raised the possibility of doing a Phase III study comparing IMRT versus passively scattered protons. This proposal arises from a recent paper presented at the annual meeting of ASTRO, wherein we compared the high-dose areas to the bladder and the rectum (the anatomic areas in which one would anticipate clinically significant toxicity from high dose radiation therapy) from IMRT versus protons (3). On the basis of that report, we think it reasonable, at least from those studies, to consider randomization of patients to IMRT and protons. Obviously there will be a higher integral dose to the pelvis with IMRT, but at least in terms of shaping the dose around the prostate, IMRT may have some advantages compared to protons (without inten-

5 Protons at the Francis H. Burr Center 65 sity modulation) delivered only by lateral fields. We think the idea might be worth testing, but the patients and their physicians would need to agree to be randomized to proton or IMRT treatment if such an effort were to succeed. For patients with locally advanced prostate cancer, who have higher PSA values and more advanced disease, we are proposing to use neo-adjuvant hormone treatment, to be followed by a combination of pelvic radiotherapy to be delivered with intensity-modulated photons to 50.4 Gray and then delivering protons via a shrinking-field technique to, initially, the prostate and seminal vesicles, and ultimately to the prostate itself. The total dose will be 82.8 CGE. Liver Cancer For hepatocellular carcinoma, Andrezj Niemierko, Ph.D., from our group, worked with Chris Willett, M.D., to design a dose-escalation study that was based upon how much dose was being delivered to the normal liver. Based upon that, these investigators developed a four-level dose-escalation scheme; Ted Hong, M.D. is currently the principal investigator on this study and we are currently at the third dose level. Protons can offer superb localization of dose in these cases. For these patients, respiratory gating is particularly important. versus 75 CGE. Pediatric Tumors A major thrust of our research effort is morbidity reduction. All of the pediatric protocols are considered morbidity-reduction protocols. The Children s Oncology Group currently allows physicians to use protons in place of photons. We currently have studies that are open for medulloblastoma, retinoblastoma, rhabdomyosarcoma, and pediatric non-rhabdomyosarcoma soft tissue and bone sarcoma. A pediatric quality-of-life study is currently open and a pediatric brain tumor study is also being proposed. The pediatric medulloblastoma protocol currently is open. Thirty-three of the targeted 45 patients have been accrued. The treatment technique has been developed and has been presented at national meetings. Currently the matchline is being done with a light field, but we are working on developing an automated device to do this, allowing patients to remain supine. A craniospinal dose display demonstrates the absence of dose to the anterior mediastinum and to the bowel structures (Fig. 1). This is a huge advance for young children. For example, girls do not have to have the ovaries removed if they are receiving proton treatment. Such a procedure would Paraspinal Tumors Herman Suit M.D., Norbert Liebsch M.D., Ph.D., and I have completed a spine/paraspinal sarcoma protocol. We treated 50 patients over about a five-year period; the study will be presented at ASTRO this year. We used protons: 77.4 Gray to the gross disease and 70.2 Gray for microscopic residuum. We used normal-tissue dose constraints 63 gray to the surface of the cord, 54 gray to the center of the cord that are higher than generally employed in usual clinical practice. Patients were planned with CT myelograms and were treated with daily diagnostic orthogonal imaging to confirm the accuracy of set-up. There was no evidence of acute or late spinal cord toxicity. All patients but one completed treatment, and that was for social reasons: it was too cold in Boston in February, so the patient decided he had to go down to Florida. Protons have for a long time been recognized as an ideal form of radiotherapy for chordoma of the base of the skull (4, 5). One can deliver total doses of up to 80 CGE or more, shaping the dose around the brain stem. Currently we have completed a randomized study of dose escalation in patients with base of skull and cervical spine chordomas and chondrosarcomas. It essentially involves patients who are either considered high risk, those patients with cervical spine chordomas or female base-of-skull chordomas being randomized to 75 or 83 CGE, or standard risk patients randomized 70 Figure 1: Craniospinal irradiation for high risk medulloblastoma. The posterior fossa receives 54 CGE; the spinal axis, 36 CGE have to be done if they were receiving photons. In the retinoblastoma study, investigators are trying to avoid radiation in these patients with the use of chemotherapy. But for patients who still have residual tumors after chemotherapy that cannot be managed with other forms of cryotherapy or heat ablation, we are using protons. The difference between photon and proton dose distributions for these tumors permits outcomes such as sparing of the contralateral orbit.

6 66 DeLaney Eugen Hug, M.D., has published a paper that examines rotating the eye medially and trying to stay off of the bony orbital structures to reduce the risk of second malignancies (6). In regard to pediatric sarcomas, we are currently treating both rhabdomyosarcoma and non-rhabdomyosarcoma patients on a protocol. The comparative dose distribution for orbital rhabdomyosarcomas is such that I think most adults would like to see their children treated with protons rather than photons. Other Protocols We have developed a protocol for partial breast treatment. It permits a patient to be treated in the course of a week, perhaps making it appropriate for someone who is busy and is trying to manage a challenging career. The proton dose distribution gives less dose to non-target structures than either a tangential partial breast technique or IMRT. Alphonse Taghian, M.D., Ph.D., completed a partial breast irradiation protocol that was presented at ASTRO in An extensive variety of other investigations are ongoing. Helen Shih, M.D., has proposed a study of IMRT versus protons for low grade gliomas; David Kirsch, M.D., Ph.D. and I are interested in using primary radiotherapy in place of radical surgery for selected patients with pelvic osteosarcomas and pelvic chondrosarcomas; Alphonse Taghian, M.D. has a treatment planning study evaluating protons for left chest wall irradiation for women with locally advanced breast cancer who have required a mastectomy and nodal and chest wall radiation; for patients who have unresected chordomas particularly for the sacral lesions we are interesting in potentially reducing the dose to the sacral nerves if we can achieve selective radiosensitization of the tumor by adding an antiangiogenesis agent; John Munzenrider, M.D., and Evangelos Gragoudas. M.D., have proposed to use an antiangiogenesis agent to reduce neovascularization secondary to high-dose proton radiation of the retina. We also are doing plan comparisons with IMRT and intensity-modulated protons. The objective is to study different planning techniques in terms of optimizing the technique for IMPT. Thomas Bortfeld, Ph.D., is leading this effort. In a related intellectual exercise, George Chen, Ph.D., has had a longstanding interest in 4D treatment planning and motion effects (7). He is collaborating with Noah Choi, M.D., and Harald Paganetti, Ph.D., studying treatment problems that arise if the target is moving in relation to the proton beam. Also in this regard, Katia Parodi, Ph.D., who had visited us from Heidelberg last year, is working with Drs. Bortfeld and Paganetti to image the positrons generated by proton interaction with nuclei. This is a potentially useful tool in situations where one might have hardware or some concern about tissue cavity interfaces and range uncertainty for verifying a treatment plan. Conclusion The Francis H. Burr Proton Therapy Center is continuing to evolve. Research is a key component of our operations. The overall intent is to identify roles that protons can play that will make a unique contribution and offer a salient therapeutic advantage. We are considering protons both as sole and as adjunctive treatment; we regard them as but another tool that we, as oncologists, can offer our patients with cancer. We believe that the unique properties of protons, notably the ability to deliver dose distributions that permit such a high degree of normal-tissue sparing and thus low morbidity, make them an especially suitable radiotherapeutic modality for combined treatment regimens. References Kozak, K. R., Smith, B. L., Adams, J., Kornmehl, E., Katz, A., Gadd, M., Specht, M., Hughes, K., Gioioso, V., Lu, H. M., Braaten, K., Recht, A., Powell, S. N., DeLaney, T. F., and Taghian, A. G. Accelerated partial-breast irradiation using proton beams: initial clinical experience. Int J Radiat Oncol Biol Phys 66, (2006). Zietman, A. L., DeSilvio, M. L., Slater, J. D., Rossi, C. J., Jr., Miller, D. W., Adams, J. A., and Shipley, W. U. Comparison of conventionaldose vs high-dose conformal radiation therapy in clinically localized adenocarcinoma of the prostate: a randomized controlled trial. JAMA 294, (2005). Trofimov, A., Nguyen, P. L., Coen, J. J., Doppke, K. P., Schneider, R. J., Adams, J. A., Bortfeld, T. R., Zietman, A. L., DeLaney, T. F., and Shipley, W. U. Radiotherapy treatment of early-stage prostate cancer with IMRT and protons: a treatment planning comparison. Int J Radiat Oncol Biol Phys. In press. Hug, E. B. Review of skull base chordomas: prognostic factors and long-term results of proton-beam radiotherapy. Neurosurg Focus 10, E11 (2001). Review. Jereczek-Fossa, B. A., Krengli, M., and Orecchia. R. Particle beam radiotherapy for head and neck tumors: radiobiological basis and clinical experience. Head Neck 28, (2006). Review. Krengli, M., Hug, E. B., Adams, J. A., Smith, A. R., Tarbell, N. J., and Munzenrider, J. E. Proton radiation therapy for retinoblastoma: comparison of various intraocular tumor locations and beam arrangements. Int J Radiat Oncol Biol Phys 61, (2005). Rietzel, E., Liu, A. K., Doppke, K. P., Wolfgang, J. A., Chen, A. B., Chen, G. T., and Choi, N. C. Design of 4D treatment planning target volumes. Int J Radiat Oncol Biol Phys 66, (2006).